Hmmm! To be clear I was reporting Canon's own MTFs as a source. While they seem to be problematic in some ways, I assumed that whatever problems their methodology had would cancel out between lenses. (For instance they don't take film/sensor into account. In fact I don't think they take production into account and instead are theoretical calculations.)
Also, the impression the Canon MTFs gave backs up my own findings with owning 2xEF 14mm/2.8, 2xTS 24/3.5, EF 24/1.4, EF 35/1.4, 50/1.0 1.2 1.4 1.8 MkI, 85/1.2 MkI, old non-USM 100Mac, 135/2 and 600/4, plus about six zooms. And a bag of Leica lenses, whole Contax G2, Mamiya 7+3 lens, couple other MFs. The experience I've had is that 1) Mamiya 7 probably beat everything, but then the 135 and 180 are so much sharper than everything there's no point in even comparing numbers. 600 could well be too but I use it at distances atmospheric haze and heatwaves always hurt the results.
As to the charts you kindly link to: I dunno! Maybe Canon MTFs aren't fit to use as toilet paper. Maybe I'm only happy with the 180 and 135 because I haven't had to buy a lens since like 2005 :-D OTOH, maybe your site's 180Mac simply sucked? If multiple sites have bad results for them, maybe they all suck, or maybe they're easy to break or something.
My last guess is that I never use it for portrait-style shooting, I'm always focusing FAIRLY close even if not macro. Maybe it gets quite blurry at long range.
Here's the Canon MTFs in question, for whatever they're worth. I'll admit the 100Mac looks awesome, but it's hardly wiping the floor with the 180. Edging it out, perhaps?
Canon's 180Mac MTF:
100mm/2.8L Mac:
And as for your claim that practically any recent EF prime is sharper, you may know best, but in that case Canon's MTFs are useless. 85/1.4L should be an easy-ish formula, not too crazy aperture, not retrofocus, and yet even at 10lp/mm, Canon's own MTFs show it's not as contrasty as the 180 at 30lp/mm for most of the image.