Motivations for studying computation and its limits:
-  Useful to know what problems are not solvable, so you don't waste your time and so you know to seek alternate formulations.
-  Sometimes hard problems are useful (e.g. Cryptography - decrypting should be a hard problem).
-  Historical / Cultural - in the last 50 years, people have spent lots of time thinking about the subject.  What did they come up with?  How did the computer come about?
Guiding principles for theory:
-  Models should be widely applicable across time and in many domains.
-  Models should be practically relevant or at least informative.
-  Build up a coherent body of knowledge over generations.