Optim said:
Summary: Current RfA ineffective. Proposed...
Your system sounds like it will produce a lot of extra work not directly related to adding content to articles. It could arguably be labelled as adding extra layers of bureaucracy. And I'm really not sure that if this system were implemented that some goodness rating being produced by an algorithm would be more useful than the present system of 10-15 people skimming through the nominee's contributions and using their eyes/brain to form an opinion.
I don't think that there's much wrong with RfA and I agree with Fred that there's too much of a conception that being a sysop is a Big Deal.
In one of the cases you mention, of user Metasquares being nominated after less than 100 edits, people have roundly rejected the nomination by c-prompt even though c-prompt knows the guy in the real world and vouches that he's trustworthy. Personally I'm all for trusting c-prompt, what happened to good faith in people and our motto of being bold? If in the (unlikely) event that things go wrong then so be it, it's not like things haven't gone wrong before.
Just my thoughts,
Andrew (Ams80)