However, they didn't really understand what makes a question valuable, or why low-quality beginner questions are low quality, and thus what exactly to expect in order to get higher quality things that could be helpful to other beginners. And yes, they did initially make some gestures towards some threshold of question difficulty, but this was a misstep (and also hard to disentangle from question complexity, as well as whether a question is interesting and to whom). And yes there were experiments like the Documentation project (which I lamented the loss of). And yes the reputation system was a totally broken idea and implementation, inconsistent with stated goals.
It turns out that when you have a critical mass of users and the alternatives are frustrating enough, you can get away with a lot of those missteps.
As for enforcement, that's what happens when the overwhelming majority of it is assigned as the community's responsibility. Misconceptions about the site's purpose compound when some users get away with them and then gain reputation from a single old/popular Q&A. (I had some, too.)
Fundamentally the site failed to figure out how to impress those values upon people before they got too involved. I think this was out of a fear of scaring users away, but clearly this was counterproductive — surfacing the misconceptions and conflicts later on only made things worse. An ideal system would have started questions closed, so that having one closed on you wouldn't feel like a final judgement (despite it not being one); and it would have shielded you from downvotes until question opening (because the entire point of this would be to recognize that questions require work by default to meet the site standards). In other worse, what the half-baked Staging Ground was trying to accomplish (and mostly did, for the small minority of questions the system decided to send there at random).