Some thoughts on Inter-Universal Teichmuller Theory

So I've been trying to think about something complicated that I don't understand, yet should at least nominally be possible.  And Mochizuki's abc proof seems like the hardest possible thing that might possibly be understandable.  After reading the abstracts, I have approximately this understanding of the proof structure.  The bold parts are what I think are the main actually interesting claims in the papers, the rest is background for my own benefit as much as yours.  This is also roughly the structure of the Fermat's Last Theorem proof (which I also don't understand), which is roughly "statements about Elliptic Curves", "Taniyama Shimura", and "proving FLT based on Taniyama Shimura".

The papers are at http://www.kurims.kyoto-u.ac.jp/~motizuki/papers-english.html , the Teichmuller papers are at the very bottom of the page.

• Part 1: A chaotic system is any system where a small change in initial input will almost always cause a large change in the iterated / Poincare recurrence of the output. [you can create your own epsilon-delta definition]. An elliptic curve is an equation of the form y2 = x3 + ax + b .   All chaotic systems are isomorphic to an extended elliptic curve.  The definition of "all chaotic systems", "isomorphic", and "extended elliptic curve" probably take up at least 50 pages each.  Presumably "semi-graphs of anabelioids, Frobenioids, the etale theta function, and log-shells" are some of these things.
• Part 2: A general method of constructing isomorphisms between chaotic systems.  This probably involves showing that both chaotic systems possess properties that are only present in one specific set of extended elliptic curves, and thus the systems must be isomorphic in other properties as well.
• Part 3: A construction of an isomorphism of the natural numbers to a different chaotic system.  For most purposes, the natural numbers (most obviously in the distribution of primes and prime powers) behave like a chaotic system.  Presumably the proof demonstrates that it actually is a chaotic system.
• Part 4: Construct a proof of abc.  Any infinite family of counterexamples to abc would probably demonstrate some type of pattern, and if the natural numbers are isomorphic to a different chaotic system where we can prove there are no infinite families of counterexamples, then there are only finitely many counterexamples to abc (which is in fact a proof of abc).