In my experience reading hundreds of IRS 990 forms these number are not out of bounds for non-profits at their scale. Healthcare and university may skew the absolute numbers, but are similar.

Specifically a great executive will personally move the fundraising needle in organizations (+1-5% or more) and mediocre executive can cause losses/missed opportunities in the opposite direction of a similar magnitude. The conventional thinking is the relative fundraising impact of a great vs a good executive may 10x their total compensation, so it's probably worth trying to retain the best.

Anecdotally, I know of a not-for-profit COO who offended an NBA player killing the relationship. His successor COO repaired the relationship, ultimately resulting in multiple millions of new donations and a co-marketing agreement with the team. When the COO was hired there were rumblings because he'd negotiated +$50k over his predecessor.

With no equity and the poor optics of commission compensation or cash-bonuses, big severance packages are one of the few ways an organization can reward employees following years of great service.

The fundraising part is the completeness of it.

I could probably COO on the non-fundraising side, but I'd be an absolute fundraising failure. Non-profit C level execs are fundraisers.

even when all the money is from a banner?

Not sure what the percentages are but many non profits live or die on a few large donors.

The banner probably drives enough small donors to give the big donors confidence that they are force multipliers and not just propping up a one-off thing.

The problem isn't that Wikimedia is paying employees too much, so much as it is that they have too many of them. Their COO compensation is not out of line for the scale of Wikimedia - But Wikimedia is way too big of an org without relation to serving their core goal.

Explorative projects like their New Editor Experience[1] were interesting, but ultimately unsuccessful moonshots. That one wrapped up (without anything to show for it, as far as I can tell). They didn't take the hint and downsize, they doubled down.

This is an incredibly common problem for organizations - "The bureaucracy is expanding to meet the needs of the expanding bureaucracy" as was put by Oscar Wilde.

[1]https://www.mediawiki.org/wiki/New_Editor_Experiences

Also there aren't many tech non profits - i imagine skillset is a little more specialized than your average non-profit.