Adminship doesn't really matter. Sure, I like my admin bit, and sadly, most of the stuff I do is directly related or implied by it, but my actual individual control over the content of this site is minimal. The amount of influence, or damage an individual administrator does to Wikipedia as a whole, or more importantly, to the real world and its inhabitants, is minute. The primary interface between the "real world" and Wikipedia is our articles - second is word of mouth, third is the press, fourth is Jimbo Wales, fifth is WMF, sixth is friends and family that are involved in the project, a distant seventh is OTRS/ArbCom/Whatever Wikipedia functionary is directly contacted by a few members of the public, and eventually, eventually an interested member of the public may contact an administrator, or whever they think is an administrator through AN or ANI. Administrators and their goings on, and for that matter, the interactions of all editors is behind the curtain. The public does not know, nor particularly care about our melodramas. Administrators are not the point source for real life problems, it is the massive size of the project and the resulting absence of quality control that is the result of our problems. This is not to say that good administration is not important to the smooth functioning of the project - but it is important to the Wikipedia community for our own sake - NOT that of the project's security as a whole, nor limiting the damage done to living persons and institutions.
We can continue hand wringing about admin standards being to high or too low, and I am all for people making their decisions on who can be trusted to be an administrator acting on behalf of the community with seriousness of purpose, but come on people. We have more things to worry about than the age and edit counts of administrators.
I worry about the model. While in the case of Aitias the RfC had fairly broad participation - and its worth noting that many of the votes against Aitias were other administrators - RfCs tend to be filled with aggreived parties and their fellows versus another party and their talk page watchers. RfCs are only useful when they increase the amount of objective perspective on a situation, and I think the format does not produce it usually. Democratic mechanisms that have low costs to create a change mean that what is created is inherently unstable. I prefer to have a stable administrator base, so I want some barrier between the democratic whim of a given sample of the community. I take as fact that those who are interested in desysopping someone are going to be disproportionately represented compared to those who are unconvinced it is necessary.