Your question is a bit confused but I think that you're mixing two types of questions and two separate policies, one of which did in fact change shortly after the question you referenced.
At the time that question was asked, all resource-recommendation questions were considered off topic and were promptly closed, by community agreement. (And, while we're here: accusations of a "small group" "taking control" of decisions like that are largely unfounded, I think, and hardly constructive.) In months that followed that thread, however, several things changed: this started with the thread
which represents a collective agreement that we should work towards a resource-recommendation policy that works better for the site, an effort which culminated with
in a tightly defined line of the kinds of recommendation questions which this community does and does not consider to work with our format and to add value to the knowledge store in our site archives.
If you have found the threads in the resource-recommendations tag useful, as they stand today, it's largely because of the effort that took place in those months to craft a policy that would allow the questions but only in a form that would make them useful to a wide audience. (In its previous life, the books tag mostly consisted of unannotated lists of books that were of little added value compared to e.g. a page of Amazon search results.)
A separate kind of question is the "how do I learn..." type of query - think this or this as examples - and those remain off-topic, though I'm not aware of a single specific thread that documents the policy.
The reason is that they are simultaneously too broad and too localized: they tend to require a much broader kind of answer than our format can really provide, and they tend to only be useful for the OP when they are strictly tailored to their experience, which makes them less useful for a wider audience. In addition, they have a junk-food-like quality to them, in that they attract lots of continuous attention (for a comparatively low objective value of the thread) and thus tend to drown out other content which we do want to incentivize.
These qualities make the genre hard to handle and it makes many people think that they're not worth the trouble, though as you note they are not entirely without interest to a nontrivial audience. A bit beyond that, the existence of Academia Stack Exchange has eased off the pressure on several of the soft-question types of thread, by providing an outlet for them - though I'm not quite sure how on-topic a question like this one, say, would be on the academia site nowadays.
Thus, if you think that those questions should be allowed, you're welcome to make a case that they do serve a part of this site's audience (which includes, according to the site charter, "active researchers, academics and students of physics) and that they can be made to work with the site's format rather than against it. An appeal for a blank cheque is unlikely to go down well, I think: instead, we would need a solid plan for dealing with the problems I listed above (and probably other objections I'm forgetting about at the moment) and a tightly defined way to handle those questions analogous to the one we currently use for resource-recommendation threads. I'm not sure that can be made to work, but you're welcome to make proposals.