It seems like it wouldn't be that hard to hide enough in the way of resources (relationships, ideas) to get a good start on Act 2. You'd lose your office and a few months of salary, and need to find new investors, but this doesn't seem as severe as "game over".
It seems like 85% of startup time is figuring out what works and what doesn't, so you can "kill" the code (i.e. shunt it out of view, and then use it as a reference in a comprehensive rewrite, possibly in a different language so there's no doubt as to your new code being yours) and your cost is only the 15% of the time necessary to implement the ideas that you now know will work.