This is a real problem. In a reading group I am currently a part of, we have imposed the following policy:
- Before showing up to the reading group each weak, each member must write a summary of the paper, like a short review, including strengths and weaknesses as well as questions for discussion.
I think that this helps ground the discussion in reality and has avoided "bikeshedding" to some extent. Specifically, writing a summary ensures that you know where the work fits into the big picture and what its main contributions are. After this is articulated, you can be sure that whatever problems you have found with the paper are in proper context (it may then be clear that they are minor problems, not related to the paper's main contribution, or it may be that they are major objections).
I think the act of finding flaws in a paper is invaluable, whether the flaws are minor or not. So the goal should not be to not find flaws, but rather to put those flaws in their proper context.
Finally, I think a good general piece of advise is to withhold judgment until reading at least the intro and conclusion in a positive light, without skepticism. That means you understand the authors' argument before you form an opinion. It may be that the negative thing you see with the paper actually has nothing to do with what it's about.