The logical conclusion to this is that ANY research is bad/faked/untrustworthy. What then are we left with when identifying truths in our world? Appeal to emotion? Appeal to authority? Picking a different method for each truth you want to identify, based on a priori beliefs about what those truths should be. Whatever- topic for another thread.
In regards to developmental norms, the typical "cut-offs" we see (e.g., "walking by 15 months"; "first words by 12 months"; "multiple word speech by 24 months") should be used as screening tools. If the behavior isn't seen by that age, the child is a few SDs from the average, and thus their MAY be a problem. Screening tools should produce a lot of false positives, as the alternative would lead to real delays not being addressed. Most kids are walking by 15 months (and most even sooner). A small proportion aren't. Of that small proportion that aren't, a portion has some has some identifying condition that is leading to their lack of independent walking, and would benefit from additional evaluation and intervention.
by having some generally agreed upon milestone age (which any well educated professional should understand is a limit on what is statistically normal and needs to be interpreted in the context of the variation in the distribution from which it is derived), we can then mete out services and resources (in this case, federally mandated EI/birth-to-three services) in a manner that ensure those who need it, get it, while also minimizing the number of children who don't need it but do get. As a society, we as a whole have identified this as the desired way of doing things, and these milestones are just a tool for achieving that ends.
As to "p-hacking" of developmental milestone research- really? These published milestones haven't changed much, if at all, over the past many years. Plus- they are something that we (lay people included) can see with our own damn eyes- most of our kids and our families/friends' kids walked starting talking by 12 mos, walked by 15 (at the latest),etc. This is a case where we have lot of naturalistic data supporting the published stuff. I find it very hard to believe that you (and I respect you) believe there are these conniving researchers faking (or judiciously evaluating) data and stats to produce purposefully incorrect conclusions regarding early development, grinning wildly as they cash their checks from "BIG EI" lobbyists. That's just adding "controversy" where there is- at best- some mild professional disagreement due to so slightly non overlapping data at the extreme ends of the distribution. I'm a lot more skeptical than the next guy. That doesn't mean that I don't believe everything I read, but rather that I have equipped myself with the tools to evaluate what I read and (in the case of a research article) understand enough about what's going on to determine if what is said in the conclusion could be and is supported by what is said in the methods and results.