It's not a judgement against people who like McMansions. It's a guide for people like me: I see a house, find it offputting, but I can't figure out why and what to look for to get a house that isn't. She gives me, a layperson, the information I need to pick a house I'll like.
Her blog actually has articles that go into detail about different architectural periods and the elements of architecture.
There's nothing wrong with liking McMansions, but I want a cute house, not too big, that I love on the outside and the inside.
Architecture is so much more complex and nuanced than that. Even "A Pattern Language" falls into the trap of providing reductive prescriptions. Ask literally any architect what they think of Christopher Alexander's actual built work, and they will either dismiss it or (more likely) won't even have heard of it.
She didn't teach me to identify that a Frankfurt house must have a wingdar, a gongle, and a shivbopped roof. It just taught me to look closer at all the different elements that make my brain say "that's a house" vs "that's a weird house"
But we can't dismiss subjective criticism on the basis that some people like the things the subjective criticism says is bad or that people uneducated in the subject don't understand. The criticism is of course not above evaluation but setting a subjective standard is not something that is impossible.
This is an important concept in software. The vast majority of decisions software developers make are subjective. Yet the vast majority of us do not have frameworks for doing subjective criticism. I think that contributes to the flame wars, cyclical fads and generally bad state of our software as code. We'd likely all be better off if we had more subjective concepts in software and not less.
[1] http://mcmansionhell.com/post/150597521816/mcmansions-101-re...