Someone who claims that faith is a good thing should not also use it as an accusation of impropriety.

I get the impression that that argument is used more to undermine claims that darwinism is a science than anything else.

Physics is a clear science; you can use the right equations and predict the motion of the Earth about the Sun, or the time a barometer will take to fall from a given height. This gives it a certain degree of credibility. The theory of evolution (and how the creationists love to remind everyone of that word, 'theory'!) is also science; but they would deny it, on the basis that accepting it suggests that it is as credible as physics or mathematics. If they insist that darwinism is a religion, then both alternatives start from the same basis of credibility; the creationists can then point out, quite accurately, that their version is older and has been around for longer, and therefore at least claim seniority.

There's a short story by Asimov that gives a very nice view of the whole argument.

Meanwhile the sociologist of religion wonders where the temples of Darwin are.

Remember that Darwinism is a lot more than biology. Sure, a computer isn't exactly an altar. That doesn't change that most of what universities are famous for in the wider world is their ideology.

Yes, but more frequently than that, the only alternative appears to be no optimization at all. Hence the heuristic.

The prior probability of someone being able to explain negative-value Will_Newsome's comments in a way that provides value for LW readers is already epsilon. Even without the disincentives.

I think that people less responding to intentionally meaningless comments is a good thing. Therefore, trivial disincentives for doing this are a good thing. Therefore, removing them in this specific situation is a bad thing.