Should the meta-MathOverflow have the Q&A format? I don't see any rationale for this. After all, meta is just an auxiliary sub-forum which should help running the main forum. Collecting the reputation points on meta seems to me simply silly. On occasions a vote can be taken to decide some issues in a democratic way. But that's not a reason for having a reputation game. For instance, if some people should be voted for some positions on MO, it should be done based on the overview of their activities on MO (possibly even in the mathematical world outside MO too, why not?) but not on reputation.
When meta follows the format of the main MO forum then instead of fixing certain MO's shortcomings it only amplifies them.
EDIT: @Emil J. has pointed to me that there are no reputation points at meta. Indeed! (Now I'll sleep better :-). Nevertheless, there is still annoying to me similarity of MO main and meta: the votes on political and philosophical issues. Here I am answering @JoonasIlmavirta's question. My answer is implied by my question https://meta.mathoverflow.net/questions/1989/mathoverflow-shape-today-and-tomorrow. I am for separating non-mathematical questions from the main MO Q&A list. One of the reasons is that voting (for and against) political & philosophical issues, and in general--voting on opinions, introduces to MO repressive censorship. It also introduces low standards, which are aliens to MO in the case of research mathematics. Indeed, a number of soft questions get a very high score, which in my opinion spoils the research character of MO reputation--instead, it becomes a popularity contest, and similar.
By the way, I love numbers, statistics and trivia like reputation points, but let ranking be objective (as much as possible) or else it is a waste.
PS. Since I have overlooked the absence of reputation on meta I am willing to remove this my Question (here go down the drain my hard-earned reputation points :-). Let me know, please.