No, I don't think it is a real problem with the GPLv2 either. The problem seems more that people take the clarifications of the GPLv3 "patent grants are explicit when distributing code", "signing bits needed at runtime are part of the corresponding sources to get the runtime binary", "the exception to the system library clause doesn't trump the separate works clause clarification", etc. as if those weren't already (implied) in GPLv2.
What seems to be happening is that instead of taking these as clarifications of the intent of v2, they are taken as some kind of fatal flaws in the old text. Instead of taking v3 and using it as a guide to the intend of v2.
Things would be much easier if people saw v3 as just a clarification of the text and intent of what v2 always already was about.