The final keyword is used for several reasons, like you said.

1. To help enforce referential transparency by disallowing assignment to pseudovariables (all fields are final).
2. To avoid what you call “think-o” bugs where you assign to a variable or argument (all method parameters and local variables final).
3. To improve encapsulation by disallowing inheritance (classes are final).
4. To facilitate closure. An inner scope (anonymous instance, for example) may not close over a variable that isn’t final.

in particular, in the examples above when you make the return value + parameters to a function ‘final’, why are you doing that?

(the general reasons for using ‘final’ i know of are: it cuts down the chances of think-o bugs; it can allow more optimizations to happen; it can be useful for ’security’ of a fashion. those general statements map down to more concrete things depending on where ‘final’ is actually used.)

thanks.