I don't think it is nitpicky. It definitely reads at least partly like an "I'm sorry you feel that I have offended you" apology and "We ... issue a public apology to Jeremy Howard..." is a weirdly evasive way of putting it. I get that it is hard to admit fault and apologise but here's what they should have said:
> We, the NumFOCUS Code of Conduct Enforcement Committee, would like to publicly apologise to Jeremy Howard for our handling of the JupyterCon 2020 reports. We did not give Jeremy a chance to defend himself before concluding that he had violated our CoC, and for that, and the stress it has caused, we are sorry.
I'm available to hire for public apology writing. :-)
Still, to be fair I don't think I've actually ever seen a public apology that was actually a full genuine public apology. This is probably the best anyone could hope for.
If you have to supply your own less-apologetic paraphrase ("sorry you feel offended"), rather than quote any such constructions in the actual text, is that really a fair reading?
Other than the apparently-magic word 'sorry', your suggested wording doesn't address most of the specific asks of ~btilly's complaint above. (Don't worry, his proposed rewording didn't, either!) Perhaps, because it's exactly tuned to what you saw as worst about his treatment, it'd adequate for you. But not necessarily for any others demanding more groveling.
It's also not consistent with the NumFOCUS committee's substantive position, which is not that they had prematurely concluded anything, but rather that they had erroneously communicated that they'd concluded something. That's a salient difference, even though the two errors look the same from Howard's perspective! (The two distinct failures would require different internal correctives, too.) You wouldn't want a public apology to lie about the actual internal errors identified just to be extra-palatable to third parties, would you?
So, don't quit your day job to specialize in public apology writing just yet!
And to entertain that idea of professionally-improved apologies a bit a more, realize that in certain situations, no matter how good the apology, or how well it hits certain notes, some on the warpath will always want X% more, & be able to find specific faults.
Short & emphatic, as per your (& ~btilly's) rewrites? "Wasn't specific enough, too patronizing."
Detailed with even a whiff of actual analysis as to the real org/psych reasons for an error? "Not empathetic enough, making excuses, rationalizing the behavior".
Long enough to include all of the emotional cloying, and the exhaustive victim's-viewpoint confession of sins, and promises for changes? "Tone-deaf, insincere because if they really believed this they couldn't have made the error in the 1st place, seems over-wordsmithed." (But also: risks setting unrealistic expectations for the future.)
> We, the NumFOCUS Code of Conduct Enforcement Committee, would like to publicly apologise to Jeremy Howard for our handling of the JupyterCon 2020 reports. We did not give Jeremy a chance to defend himself before concluding that he had violated our CoC, and for that, and the stress it has caused, we are sorry.
I'm available to hire for public apology writing. :-)
Still, to be fair I don't think I've actually ever seen a public apology that was actually a full genuine public apology. This is probably the best anyone could hope for.