OOP is a good idea some of the time, and Java's OOP is good enough a lot of the time. The problem I have with Java is that the programmer must use OOP in all sorts of situations in which it isn't really the right thing to do to make up for deficiencies in the language. Particularly, the language does not have first-class functions or closures, and the boilerplate necessary to fake these features is so big that I would seriously consider solving a problem in a worse way.
The danger is that a person may think that the way things must be done in Java is actually a worthwhile way of doing things in general. Such a person would be in a state of permanent confusion in which the meaning of the question "How will I solve the problem?" has become "How will I partition the behaviors and data involved in a solution to the problem?" This second question is a good thing to think about, but not before the first one, and certainly not instead of the first one.
I don't think it was a dumb question. I didn't actually encounter any people like this for a long time.