Is It Reasonable for a PhD Student to Worry about being caught in the academic version of a 'Roach Motel'?

Yes, according to this summary of a recent paper by Paul Oyer of Stanford University. See the discussion of this paper at Slate.

If you are concerned about ending up at an institution that "you cannot check out of", then the advice is straightforward --- go to the best graduate school that you can get into that will pay your full way, decide on a research program early and work closely with the top person at your school in that field, write early and often, attend conferences, and publish (and the higher the placement in terms of rank of journal the better).

Good luck.