I think first and foremost, you need to supply enough context for it to be a question. As-is, it's an open-ended family of questions. Basically, you could think of every language to which you might compare C++ as a plane. That plane has tasks defining one dimension, and programmers with differing backgrounds (and, to some extent, comfort with various abstractions and such) in the other dimension.
As I mentioned in my comment, it also depends a great deal upon what you mean by "learn" (for one example) -- whether you mean learning enough to use the language reasonably productively, or language lawyer-level expertise (and believe me, in the case of C++ those aren't even remotely similar.
That leaves us with an essentially infinite volume, with each point in the volume defining a single question to which there's a reasonable-sized set of answers that stand a chance of being meaningful (even if they are, to a least some extent, still subjective).
I suspect, however, that you'll run into a bit of a problem: once you've really defined the parameters of the question, you'll run into one of two things: either you've defined a comparison that really does fit the Gorilla vs. Shark description @Mark Trapp gave, or else you'll have defined a comparison for which the answer is almost blindingly obvious. There may be a few for which a meaningful, thoughtful comparison is possible, but my initial guess is that they will be few and far between.