It would help me to understand why my version is a straw man if you would steel-man it.

There isn't a convenient steel man available. Not all wrong (or, to be agnostic with respect to the correctness of our positions, disagreed with) positions have another position nearby in concept space that is agreed with (or, sometimes, disagreed with only with significant respect and more complicated reasoning).

As it stands, I am puzzled by your accusation because Eric Raymond said, "Let’s drive away people unwilling to adopt that 'git’r'done' attitude with withering scorn ...". Why is it a straw man to characterize this as "heaping scorn on people and seeing who sticks around"?

Because that is a different described procedure. They are similar in as much as scorn is applied in both cases but the selection process for when scorn is applied is removed and the intended outcome is changed.

To illustrate, consider taking the required equivocation back in the other direction. We end up with:

Empirically, <driving away people unwilling to adopt that “git’r'done” attitude with withering scorn, rather than waste our time pacifying tender-minded ninnies and grievance collectors> leads to lots of time wasted on flame wars.

This seems to be a different empirical claim. It is also a more controversial claim and one that is less obviously correct. I certainly wouldn't expect scorn to be the optimal response in such circumstances but the claim that it wastes more time than the described alternative is still an empirical claim that would actually require empiricism to be done and cited. It isn't something that I have seen anywhere.