As for too much time on his hands - yes, he admits as much on the site!
I mean, you have a theory, and you take the subsequent data and try to make it fit the theory ... that's the real question. Since he came up with his theory about what LOST is in 2006, is all the rest just trying to make everything that has happened since "fit"?
If the guy is right, hats off to him.
Edit: thinking about that, it would give the writers so much latitude, so much opportunity to set / reset, recompile ... what a dream, you don't have to follow the rules when the rules are more widely spaced than just the parameters of reality.