I think your skepticism is perfectly reasonable. I'd first observe that humans tend to be greatly influenced by tradition, and tend to look for existing models to adapt to new endeavors. Mission Control has a structure that's clearly adapted from seafaring traditions--just map the job titles: Flight Director = Captain, GIDO = helmsman, FIDO and the backroom boys constitute the Chief Engineer and his crew, Flight Surgeon = Ship's Doctor, etc etc etc. It seems as though all those people who founded MCC who came from aeronautical flight-test backgrounds, nevertheless instinctively looked at naval operational models for inspiration.

You gave several examples of leaner operations, and they all come from aeronautical traditions rather than naval. Still, I have to point out that you left out the support customarily provided to military flight crews on missions. Isn't an AWACS another form of detached Mission Control? The dividing lines are a bit different and the individual aircraft have more autonomy than spacecraft, but we still detach and centralize functions such as theatre radar and communications. Same with civilian aircraft; maritime vehicles are probably the best examples of nearly-autonomous transport we have at the moment.

I think you're probably right that copying naval models results in a somewhat featherbedded Mission Control. The right model is probably a synthesis of the two traditions, but don't forget that a new model can't be so new that people find it artificial and hard to adapt to. What might make sense from a strictly engineering point of view might confuse or disorient social primates. Whatever it is, it will have to feel something like a tribe, or humans will reject it.

"Of course, this doesn't count maintenance personnel between missions." is a pretty big "of course". Maintenance personnel would have to be carried aboard an autonomous spacecraft. Even in our solar system, transit times will be measured in months for the foreseeable future, so drydock won't be a feasible option when something life-/mission-threatening breaks. Multidecadal interstellar missions? Better bring not just the garage but the factory too, and enough raw material to reproduce anything that breaks.

Seems to me that the maintenance and self-repair factor tips the balance further in favor of human-crewed autonomous spacecraft. Consider it a given that whether the ship is robotic or crewed, it will require the same amount of maintenance and have the same probability of failure (ack'd that life-support and other uniquely biological infrastructure does increase complexity and risk of failure, it's probably not greater than a factor of two, ignorable for back-of-the-envelope noodling).

On a robotic craft, self-repair means the addition of a separate infrastructure, with separate supervising AIs to maintain independent cross-checking capability, and specialized machinery such as mobile robots carried along but held in reserve until needed. Self-repair on a robotic craft is a pure burden.

Humans provide the economy of flexibility. In a human crew, a few specialists would be devoted to maintaining the infrastructure, but most of the crew would attend to other specialties on a day-by-day basis. In the event of an emergency, a properly cross-trained crew can instantly be repurposed to repair the damage. Machines specialize, humans generalize. You get the equivalent of self-monitoring AIs and robotic repair crews effectively for free with a human crew.

Would the cost-savings justify the extra biological infrastructure? Maybe, maybe not...but people prefer the answer that supports sending humans, so I'm happy to throw another log on that fire. I'd prefer to send humans too. I never claimed I didn't have a horse in this race.