Hypothesis: Google data indicates that pseudonymous users make a net-negative contribution to the quality of an online community (c.f. The WELL). They want a real name not for any commercial reason, but to hold users accountable for their behaviour. Google's real names policy for G+ isn't part of some complex conspiracy, but in fact standard Google practice - following the data, even when the data contradicts what users say they want.

Is it evil to put the interests of the entire network above the interests of a small minority? If you can say for sure, you're either an ideologue or you've solved the most fundamental problem of politics.

Personally, I think there's room on the spectrum for a variety of standards of identity. At one end there's 4chan, at the other carefully-vetted private communities. In the middle you've got sites that require a "real" e-mail address, or Something Awful charging $10 to join. Google's real names policy is distinctly flawed in execution, I think largely as a result of scale, but I don't think it's a fundamentally bad idea and certainly not categorically evil.