Bob Crawford writes:

How do you substantiate your initial claim that "[Real life GA]
autopilots make coordinated turns even when they cannot control the
rudder"?

By watching the ball in the turn indicator. It moves far less in an AP turn
without rudder than it moves when I make a turn without rudder. It leads me
to believe that the AP rolls the aircraft in a specific way that minimizes
uncoordinated flight. I'm trying to figure out the best way to make a turn
without rudder in a similar way, since I figure it will help make coordinated
turns _with_ the rudder.

The only proof of this supposed phenomenon you've provided is that it
occurs in MSFS.

MSFS is a simulator. If it happens in the sim, it probably happens in real
life.

Additionally you've offered no proof that MSFS models autopilot
behaviour correctly in this respect.

I have no reason to believe that it is incorrect. In fact, Viperdoc's
description of the real aircraft matches the behavior in the sim.

With that in mind, and given the statements from real life pilots that
real life autopilots do not behave as you describe the MSFS autopilot
doing ...

See above.

... a poorly modelled rudder / yaw response or AP seems like a
perfectly logical conclusion.

It's an attractive conclusion (for those who wish to disdain simulation), but
not a logical one.