I just had this idea pop out of nowhere in my mind for a puzzle. The concept is of an icosahedron that spins around corners and that can jumble. The main thing that i find would be a problem in the creation of this puzzle is how to make it jumble. It is something that i have trouble understanding, but from what i've seen it would seem like puzzles jumble better when not pillowed and do not have curved cuts between each piece.
I have seen some other icosahedron puzzles on the market, take the Tony Fisher's one as an example. 20 sides indeed, but a corner turning icosahedron with a mech similiar to a dino cube seemed like an original idea to me. And of course as we know puzzles that turn around corners are (generally) easier than face or edge turning ones, so i thought that jumbling would have been a nice thing that could have been added to this vague hipotesis for a puzzle. Could such a thing be made?

You cannot force a puzzle to jumble. It is just a natural property of a shape and it doesn't matter if it is pillowed or uses curvy cuts or not. It's not something you plan to design for. It's just a natural consequence. The original builders of the helicopter cube (or bevel cube) didn't realize that it could jumble when they designed it, but while playing with it, they found the jumbling move. Similarly, someone had to point out to bhearn that face turning icosahedra jumble before he realized that the puzzle he design and owned for several months could jumble.

A corner turning icosahedron is equivalent to a face turning dodecahedron. Face turning dodecahedra do not jumble.

That just left me totally speechless. I understand. I just thought that an icosahedron that jumbles would make for a good puzzle. Not corner turning, i guess...

I've known about that for a while, but I have a question: how come the Dino-Dodecahedron doesn't jumble? Is it even possible to change it into a FTI?

It's not cut deep enough to jumble. A trivial example of this is with the very-shallow helicopter cube:


The same thing happens with FTI's; the jumbling on a deeper cut FTI happens between faces that commute on a Dino Dodecahedron but don't on the FTI.

Is then there a way to create a jumbling puzzle that rotates around the corners?

Of course there is. You can simply take the Dodecahedral version of an FTI and it will jumble and turn around corners just fine.

I would assume such a puzzle has been already made, due to the shear amount of things shapeways produces. If it hasn't been made, though, i think it would be a very cool puzzle.

http://www.shapeways.com/model/300709/

That thing is definitely awesome. I really like dodecahedra and jumbling puzzles. That's like a mix of the two. It's a tad expensive for me though...

Do you like curvy cuts? because if you do there is a puzzle that's already prototyped by mf8 that would be made just for you.
twistypuzzles.com/forum/viewtopic.php?f=1&t=22236
You gotta scroll down a bit. Rox has it in her hand. There's a jumbled picture too. Chances are you will be able to obtain one in the next couple of months.
(I did about this exact post a couple of days or weeks ago somewhere already. How do you prepare a link so that it shows the correct line of a page? )

That actually looks pretty damn cool. I never had an mf8 product before so can't say much about quality but that really seems like a great fun puzzle.