Science is now, basically, dead (my direct experience of science is inevitably partial - but the same mechanisms seems to have been at work everywhere; even outside of medicine in the humanities some of which I know reasonably well - and the social sciences were essentially corrupt from the word go).
What we think of as science is now merely a branch of the bureaucracy. It would, indeed it does, function perfectly well without doing any useful and valid science at all.
Indeed, modern professional science functions perfectly well while, in fact, *destroying* useful and valid science and replacing it with either rubbish or actively harmful stuff (this is very clear in psychiatry).
I find that I now cannot trust the medical research literature _at all_. I trust a few individual individuals but I do not trust journals, not fields, not funding agencies, not scholarly societies (like the Royal Society, universities, or the NAS) not citations, not prizes (Nobel etc) - in my opinion, none of these are trustworthy indices of scientific validity - not even 'on average'.
The system is *so* corrupt that finding useful and valid science (and, of course, there is some) is like finding a needle in a haystack.
The vast bulk of published work is either hyped triviality (which is time wasting at best), or dishonest in a range of ways from actual invention of data down to deliberately selective publication, or else incompetent in the worst sense - the sense that the researchers lack knowledge, experience and even interest in the problems which they are pretending to solve.
So, what should a person do who wants to do real science in an area? - if (as I think its probably the case) they need to _ignore_ the mainstream published literature as doing more harm than good.
Essentially it is a matter of going back to pre-professional science, and trying to recreate trust based interpersonal networks ('invisible colleges') of truthful, dedicated amateurs; and accepting that the pace of science will be *slow*.
I've been reading Erwin Chargaff lately, and he made clear that the pace of science really is slow. I mean with significant increments coming at gaps of several years - something like one step a decade or so, if you are lucky. And if 'science' seems fast, then that is because it is not science!
This is why science is destroyed by professionalism and its vast expansion - there are too few steps of progress, and too few people ever make these steps. Most 'scientists' (nowadays in excess of 99 percent of them) - if judged correctly - are complete and utter failures, or indeed saboteurs!
So science inevitably ought to be done as a serious hobby/ pastime paid for by some other economic activity (which has usually teaching, but was often medicine up to the early 20th century, and before that being a priest).
Why should anyone take any notice of these putative small and self selected groups of hobby-scientists? Well, presumably if they produce useful results (useful as judged by common sense criteria - like relieving pain or reversing the predictable natural history of a disease), and if the members of the group are honest and trustworthy. But whether this will happen depends on much else - their work may be swamped by public relations.
So, groups of practitioners are best able to function as amateur scientists, since they can implement their own findings, with a chance that their effectivceness might be noticed. And in the past groups of practicing physicians would function as the scientists for their area of interest.
This seems the best model I can think of for those wanting to do science. But science is intrinsically a social activity, not an individual activity. So if you cannot find or create a group that you can trust (and whose competence you trust) - then you cannot do real science.