A half-rationalist that believes he's not a complete rationalist holds off on investing in the alchemy project, and attempts to figure out whether their mind is doing everything correctly.

Preferably, a good half-rationalist is prioritizing becoming a complete one before having to make any important decisions or ending up with a corrupt anti-epistemological belief network. They would be aware that they can't fully trust their brain.

A half-rationalist that believes they believe the above, but don't actually update on this...

Well, let's just say my brain isn't quite focusing the picture all too clearly, and I can't tell what would happen in this set of examples. My question in the grandparent was mostly equivalent to "Is this last example going to be worse than the one who invests it all in alchemy, or slightly better? My first guess is it's going to go even more horribly wrong."