I agree with your formula Amanda, but it feeds on itself if you include rank as a factor. More selective = higher numbers = higher rank = employer interest = better job prospects = more selective = ...
I think as long as rank is a factor it will distort the evaluation of how good a school really is. Unless selectivity is removed from consideration. Maybe there could be two categories of rankings. One would be for quality of education one receives and the other for job prospects upon graduation. Or maybe job prospect data could be provided without factoring it into rank. I think if that kind of ranking system became popular, it would be a useful tool for employers who could be better assured of the quality of education of their new hires. Isn't that better than the current system, which is based on prestige and elitism? They just might hire more often from, say, Tulane than Georgetown (just a hypothetical example). Vermont and San Diego would get a deserved boost.
If you take out selectivity, you're left with how good of an education you get, for what that's worth. There seems to be a market for brand-name JDs. Maybe under an alternate ranking system the market would favor a quality education.
Also, I assumed this given: selectivity is one of the more manipulable factors of rank. Schools play with the numbers in order to move up and it's not actually related in itself to the quality of education the school provides. That's why it's inherently flawed as a factor in rank.