Math kind of came crashing down today; I'm back to where I started when I first tried revising the (apparently successful still) first completely constructive (but still aesthetically imperfect) proof I came up with. That is, all of my aesthetic improvements have failed, but the basic proof still seems correct. Well, one aesthetic improvement worked; it was just capturing a whole bunch of inductive lemmas in one, überlemma that quantifies over pairs of second-order propositional functions. You heard me right. It's essentially a function of type ((prop → prop) → prop) * ((prop → prop) → prop) * (prop → prop) → bool that I prove to be constantly true under the right assumptions.
Went to D's in the evening. Some people were actually there! It was pleasant to talk to them.