Well, I would suggest that the rankings you are looking at are a bit more arbitrary than the numbers you are ascribing to them, and that anything in the top 20ish or so is really in the same grouping (eg is Duke really "better" than Cornell? Probably not -- it turns on NIH research funding, which means squat to a med student), so you really have more like 15 in one group and 5 outside. That may pan out fine, but to actually be applying broadly you probably would want to move a few from the first group to the middle ones. That's what applying broadly means. Doesn't mean everybody with every set of stats needs to apply broadly, just that you aren't.
One caveat though, it's not all about the numbers, and EVERY YEAR someone on here with your stats gets burned because they almost exclusively apply to the top 20ish schools, and are not deemed a "good fit" by those schools, while a more broad grouping of schools might include some who would give more deference to the high numbers and overlook some of the less strong aspects of an application. You have to realize that after a certain numerical point at the top schools, the law of diminishing returns kicks in, and so just because you have that 3.9/39, doesn't mean the school isn't going to be more wowed by the former marine mountain climbing kickboxer with the 3.6/36 and the competitive violinist with the 3.7/37. So I'd have to say that once you have slightly higher numbers than the school's average, it starts to matter less than the other things.
So whether you have applied broadly enough turns on the strength of what else you have in your app. If you have the high numbers but nothing more than very generic hospital volunteering or research, I'd say add more middle ground schools. If you are as accomplished in the non-academic stuff than the academic, then I'd say, sure, focus on the "top 10", or "top 25" or wherever you are arbitrarilly drawing a line. But don't kid yourself into thinking that that's applying broadly.
.
