I'm super quanty, so I imagine if methodological training is your concern, my preferences are fairly in line with yours. I just think that with very few exceptions, no school offers EVERYTHING one needs in the way of methodological training, which is why the summer institutes (IQMR, ICPSR, EITM, Networks Conference) are so useful.
Jasjeet Sekhon went to Cornell and he's currently the quant meth ninja of the universe. If you want more methods training, that's what ICPSR is for... and you can always audit classes in the statistics department.