I'm so relieved, happy and delighted you made it, even though I personally was not able to spare the cash. Congratulations forever! :-) I'll be looking forward to the final game so much.
At the previous, relatively old post by Eternal Gamer, yes, I'm afraid your definition of an accessible app *is* in deed a bit too super strict. Screen readers were not designed to support games. This game, like many other modern mainstream ones, is made in Unity, which doesn't compose its GUI's using the standard graphics API's and controls provided by the system but rather through custom sprites that you as the developer later fill with any textures of your choice, if I understand it correctly, and there really is no way to make the current screen readers able to support that setup, so I see it as a minor miracle that an Unity dev implemented at least Sapi support, and, most importantly and above all, was motivated enough to do that in the first place. Just look at the screen reader support problems many devs face even when creating audiogames as such. If someone wants to make their game output text directly through a popular screen reader, they will usually come across one or another programmatic issue regarding the obsolete and relatively unfriendly, closed setup of thescreen readers sooner or later. Let's face it, even though screen readers should be providing us with access to as much information and as many different kinds of GUI controls and applications as possible, the amount of information they expose through their own API's, and especially the ways in which you can control that provided information, is pretty limited, so much that it's not sufficient at all for a third party app to be able to hook into the screen reader efficiently. And, last but not least, games just inflict new controls, concepts and paradigms at you as the player in general anyways, so I don't see any problems with that myself.
I mean, as an end user, I would tend to agree with you wholeheartedly. Unfortunately, though, we are not yet living in an ideal world and the amounts and complexity of the technical obstacles currently preventing such an ideal universal access support solution is still too large. However, I see other areas where this problem matters much more than with games, since as I said, games do expect you to learn new controls and ways of thinking anyway, so what.
Just my thoughts. I'm not sure whether there'd be any good in this evolving into a discussion though. Quite a lot of such discussions have already been held in the past, I know, and they never led to anything as the screen reader developers won't just be making as complex a redesign of their software as would be needed to make that kind of support possible, after so many years spent coming up with the relatively solid and stable code bases they already have. Asking them to do something like that would be almost like asking, say, the developers of Twitter to integrate Facebook, Skype and Whatsapp support all at once into all their native websites and clients. :-D