>
One major change I would make from public money is publishing of results good or bad. There's this competition to get into top tier journals (which I don't think is bad or you can get rid of), but this is done at the cost of dumping good science just because it wasn't good enough for a top tier. Using impact as a metric is good but not great.> The mark of a good scientist is their work and their passion, not where they publish.
You are overlooking an invisible but equally important mark of a good scientist: a scientist’s impression on peers aka reputation, which usually comes down to two types: prestigious publication (in a prestigious journal or with a prestigious co-author) or having a good number of citations. The first kind of reputation is bequeathed almost immediately, the second kind of reputation is deferred — it can only be gained after the passage of time (years).
If a scientist is known to have co-authored a paper that peers later on characterize as a seminal contribution to the field, peers are much more likely to pay attention to future papers from that scientist. The converse is true.
The number of published papers, whether in prestigious journals or otherwise, far exceeds the number of people with free time and relevant qualifications to assess them.
At the end of the day, human attention is a scarce resource, which is why it is the metric those aspiring to be regarded as “good scientists” in their field optimize for. Taken together, a scientist’s publication history is used as a proxy for authority in a field, which in turn is used to filter which papers to read or send to the trash can.