Most of the time this is true, but I've certainly had comments like this one[1] where I thought maybe deserved some upvotes, but an order of magnitude more than what I would consider other good comments. And sometimes I post what I think is a wonderfully insightful post about how simplified memory access methods are the most important part of RISC nowadays and nobody upvotes it.

[1]http://news.ycombinator.com/item?id=4789008