How is administrator attrition being addressed?

Respectfully, I don't think your reasons are completely valid. Projects value users who are more active because they do more work. To go through your bullets one by one:

1. Those who pass RfA are commonly unknown to half the people who vote for or against them. These people are evaluating their edit history and answers to questions. Being "known" will bring in people to vote for (or against you), but nobody votes against someone just because they don't know them.
2. I don't think this is true. There's too much going on with Wikipedia to know every outcome of every discussion, and everyone knows this. Everyone who has enough experience to be an admin knows this.
3. I think you're speculating here.

I think it is important that we not conflate inexperience with inactivity. We shouldn't be giving admin tools to those with no experience, but there's no reason to keep them from an editor who has an impeccable two year record, even if he only edits twice a month. That editor will do some good with the admin tools, and if we had a thousand more like him, there would be a lot less load on the more active ones, which means they can spend more time writing/editing articles, innovating, or participating in areas that they most enjoy.