This puzzle deserves a better recognition!
As there has been a single reply after my post in Timur's thread, I have decided to try it again with a new thread.
I have edited it at a few places only. An important addition you'll find under EDIT at the very end of this post.
So, if you have read it in the other thread already, you may want to go to the end of this post directly.
I hope to reach some more readers, though

Actually I have missed Timur's thread until last week. Timur has wondered already, why I have not made any comments.
For the impatient reader: “A long story has come to a happy end! I can certainly recommend this puzzle.”
Here follows a review of the Master Face Turning Octahedron.

History
Timur has posted originally here. As you may remember, the original puzzle was a mixture of casted and printed pieces.
Here are a few quotes from this thread:

I could go on for a while
And now it is available at Shapeways at a fair price, in a better quality, and nobody wants it? (besides me, who has one already )

I have won the prototype on Ebay on November 24th, 2010.
I have posted to the rare puzzle thread this on December 1st. Quotes from there:
Timur had written: “If a turn seems apparently hindered, never apply force. In case you do, you’re not likely to break anything, but chances are you’ll see the mechanism in all its beauty."
I had made that experience and even not applying any force.
“My verdict is: Timur has not exaggerated and my feeling of danger is quite justified.
I'm afraid I'll never be able to solve it, because I will see "the mechanism in all its beauty " too often.
Definitely, a printed version would be a partial solution, only. IMHO, some real enhancements of the mechanism are necessary to make it more stable.”

I have written a PM to Timur asking, if he could make an improved fully 3D printed version on December 5th. As he has said above, if the buyer (me) does not care about the aspect of uniqueness, nobody can blame him (the designer), that the (pretty high priced) Ebay puzzle is no longer unique. I liked the idea that I could influence Timur that he would make this puzzle available in an usable and playable form.
I got PM from Timur on April 5th that I could buy it at Shapeways (I had missed this thread from March 29th). He made a special, unique price for me – as the unique Prototype buyer.
There was some delay at Shapeways and I have received the parts on April 21st. Unfortunately, Timur was on vacation at that time. I got some important hints for the assembly yesterday. And now I have another unique version of this puzzle (Unique in the sense, that I have been the first one to order it on Shapeways. I hope, I'll be not the last one.) Let me state here, that I do not like so much if puzzles are available to a single person at a unique price. I'm more for the democratic approach supported by Shapeways for custom puzzles.
Until yesterday I lived with the impression that the improved version is still a secret

Assembly
I have ordered the puzzle in BSF (Black Strong Flexible, which is White Strong Flexible dyed black).
Next time I'll try WSF and dye the parts with a black marker pen.

Actually, I needed a fundamental - yet simple - hint from Timur: Start with the ball core and adjust the screws until the turning feels right. The outer parts can be installed afterwards.
I had made the mistake to make the screws completely tight and that caused problems.
So, make the screws a bit loose at the beginning and assemble the 14 pieces plus the four-armed spider, making the ball core with the six corners extruding. (BTW; no washers and springs needed, just 4 M3 screws 10-13 mm long):
- The four triangles in the picture above
- The six corners
- The four triangles
Actually, with these simple hints things are quite obvious.

The rest is not very complicated as well:
- The 8 centres go with two types of adjacent outer centres, four faces with larger triangles
- and four faces with smaller triangles

This makes two types of faces which are opposite to each other.
Build large edges that consist of two edge pieces and two of those outer centre triangles of different type and you will find out easily how things fit together.
Don't be afraid, with these hints the assembly should be pretty straight forward.
One more hint: Some pieces (the little triangles) have tiny holes where superfluous powder could have remained.
Knock them on a table to get rid of this powder

Turning
The puzzles feels a bit loose, but pretty stable. No dangerous feeling at all. It turned quite smoothly right after the assembly.
It is a great relief that it turns so much better than the prototype.
Honestly, I have not solved that one and I'm not one who is complaining easily about badly turning puzzles
Still, you have to align it and it is not for speedsolving. But you will not have expected that from such a complex custom puzzle, right?
All in all, I'm quite pleased how it turns.
It will certainly improve over time as all my Shapeway puzzles. A few more sessions of lubrication (if you are familiar with WSF, you know what I'm speaking about) and some more breaking in and it will be excellent.

Please, look at the EDIT below for an important addition!

Stickers
I have sent a Coreldraw template to Olivér (Nagy) of this forum and he has produced nice stickers for the puzzle. I followed Timur's advice and made the stickers with an edge length of 12mm before rounding the corners.
I'm quite pleased with the result. Here are some views of the stickered puzzle:

Olivér has lots of colours to choose from.
He offers a set for € 4 plus shipping.
I have used Greg's ironing method in two steps:
- Ironing the surface of the unstickered puzzle right after assembly.
- Ironing on top of the stickers with less heat and just for a few seconds.
It has worked quite nicely (A funny observation: One colour - purple - did not adhere well. I had two sets and both
didn't work. I have replaced the purple you can see on the picture by dark blue and everything is fine.

Size
It is a bit smaller than the (unique = I'm the only one who can make that comparison ) prototype and has the same size as an FTO from Taiwan:




Solving
I had developed the necessary algos on Gelatinbrain 4.1.5 back in December.
Actually, I managed only one single commutator once on the physical puzzle and gave up.
That was the point in time where I have asked Timur, if he could make a better version.
I have never solved a scrambled 4.1.5 (I have never tried any scrambled puzzles on Gelatinbrain) and I'll have to wait for another six weeks, because this master piece will be my birthday gift
My impression is, that it is in the same category (how hard it is to solve), as the Pentultimate or Starminx V1 (the real V1).
Maybe the Gelatinbrain experts can compare it better?

Conclusion
I would be surprised if this will ever be mass-produced. At $ 250 it is a very good Shapeways puzzle.
If you can afford it, I wouldn't mind if you go and set and end to the state of "uniqueness" of my puzzle.
Whoever wanted to beat me on Ebay should order it immediately and get a much better puzzle for much less money

EDIT: After I had made my first reply in Timur's thread and before it got packed away as my birthday gift, I tried some algorithms. As always with custom puzzles, the puzzle was a bit harder to turn and to align when doing real life sequences compared with random breaking in moves. I have had two pops and larger parts of the puzzle came apart. So, either you handle this puzzle with care - as I do it now after my "accidents" - or you will be very well trained reassembling it, after a while. The good news: It is completely unlikely that you will break any pieces.

I do want it, it's a matter of being able to afford it. But the puzzle is certainly one of the most beautiful puzzles available on Shapeways, and I may start saving up for one.

I think it looks like a fantastic puzzle! I just need to get some money, a face turning octahedron, some more money, and then this! I hope to own it one day.

By the way, I do appreciate these reviews - with so many puzzles being made, it is hard to narrow down decisions, and things like this help a lot to sway me one way or another, so thank you!

Chris

It is definitely not that I don't want to own some of these fantastic puzzles. It is that I just cant afford to spend $200+ let alone $100+ on a puzzle. With so many companies mass producing peoples designs it would be great to get someone like MF8 on board with peoples custom puzzles. MF8 has been replying to my PMs about getting certain puzzles mass produced.

I've got this via PM from Katja. As you can see, I have her agreement to quote it here:
I think it is related to my review, and I wanted to share this opinion with you.

I think that is very difficult to compare the difficulty of puzzles. I mean some difficulty levels are quite obvious: A 2x2x1 is certainly much easier than a 2x2x2. Comparing a Starminx V1 (face turning) and a Pentultimate is not so easy. Number of piece types and number of different algorithms needed can be a hint, but can be misleading.

What do others think?
(Especially, how would you see the difficulty of Gelatinbrain 4.1.5, as the physical puzzle is not very whitespread nowadays ?)

Hi Konrad,
I'm surprised that you find the Master FTO difficult. Considering you're not a Gelatinbrain solver I've always been impressed with your solving ability. I agree that it is difficult to objectively compare puzzle difficulty. Even though you can solve the Starminx points with [3,1] commutators it is very hard to find the sequence.

When I first solved all of the big FTOs on Gelatinbrain I thought they were somewhat difficult. I didn't have a lot of solving experience at that time though. I just re-visited 4.1.5 and I quickly and easily found short, pure [3,1] sequences for the pieces. There is a 12-move 2-2 swap for the face centers but I think you should solve corners and then centers so you don't need any cycles for the centers at all.

There is no question in my mind, the Starminx is much harder than the Master FTO. If you'd like me to list routines I'd be happy to.

Thank you for your kind words, Brandon. Actually, I like Gelatinbrain very much as a help for finding algorithms. There are two reasons why I have never tried to solve a scrambled Gelatinbrain puzzle:
- I find pattern recognition much harder on a computer screen showing the puzzle in two different views.
- My wife thinks (and she is right ) that I spend too much time on my computer and with puzzles, already.
I deeply admire the solving abilities of you and e.g. Katja, Elwyn, Julian, schuma ...
In my review I have tried to express my expectation how hard the Master FTO is compared with other puzzles, I had solved. I wrote: "I had developed the necessary algos on Gelatinbrain 4.1.5 back in December.
Actually, I managed only one single commutator once on the physical puzzle and gave up."
That means, that I have all necessary sequences for the Master FTO, but I have never scrambled and solved one, because the prototype was so hard to turn and the new one is my birthday gift
My current situation regarding the Master FTO is quite similar to the situation I had back in January regarding the Starminx: I have the necessary algorithms, but no turnable, physical puzzle. Solving a puzzle has several main aspects for me:

- Finding the algorithms. I do not recollect the details, but I have searched for the algorithms around the same time in the following order: Pentultimate, Master FTO, Starminx V1. Originally I had a quite lenghty procedure for the Pentultimate corners (long but no really far fetched) and then I digged up a 10 move sequence from you in the Gelatinbrain thread. If my memory doesn't mislead me, I recollect this "finding of the algorithm phase" not much different for the three. That does not refer to your 10 move sequence, but the long and easier. I still wonder , how you had found it. (I guess with the help of a computer?)

- Turning quality of the physical puzzle. My concentration suffers a lot when I'm getting locks and pops. This was so bad on the physical Master FTO, that I have never scrambled and solved it.

- The confusing (or dazzling) factor while doing setup moves and remember them reversing them correctly. The Master FTO is certainly easier than the others regarding this aspect.

On the other hand, I look at the number of piece types:
- Pentultimate: centres (12), corners (20)
- Starminx: centres (12), edges (30),star tips (60)
- Master FTO: centres (8), outer centres (24) corners (6), triangles adjacent to the corners (24), outer edges (24)
The two more piece types of the MFTO seem to compensate somehow for the lower total number of pieces, compared with the Starminx.
Thank you, but...
EDIT 11/06/16: I have solved it (after I have got it as my birthday present) and I agree: It is a bit easier as the Starminx.