I think the plug connections is the problem. I think you have intermittent connectivity in that plug. Sometimes it's fine, sometimes it's not. When it's not, the engine reverts to limp home mode where it just uses the MAP sensor and TPS sensor to guestimate the air flow, and it your case it looks like it does a pretty good job of limping. Better'n Festus, anyway.

The internal guts of the MAF sensor is all solid state, and not nearly as likely to fail as the connector is to get either "almost" plugged in, or a tad corroded, or some other 'lectron gremlin that keeps the little 'lectrons from going where they're supposed to. Might possibly have some air flow crud on either the ultrasound emitter or receptor in the Karmann vortex sensor, but they are after the air filter.

My thought on too rich is a stuttering connection to the MAF that's throwing off the measured voltage that the maf sends to the ecu, in effect, telling the ecu a wrong number, in this case showing more air than actually passes to the motor, making the ecu dump a very rich mixture.

I don't remember more than a very few cases of actually bad MAF's on here.

The reason I don't think it's just a "going to open loop" problem is that it shouldn't go real rich in open loop if the maf is doin' right and you have consistent ignition and the tps and map are also ok, as is the rpm signal and the cts. Closed loop has a pretty limited amount of trim factor compared to the total afr in this kind of system.

The intake air temp sensor is also in the maf housing and uses the same connector, and may also be sending wonky for the same reason. The ecu uses the map and the iat sensors for readings for mixture corrections for varying ambient conditions. If the iat is sending an incorrect voltage over time (it's a thermistor, and varies resistance with temps to vary the voltage signal the ecu hears), you might get and incorrect air density, also throwing off the ecu calc'ed fuel amount to rich. I'd have to logic some more on that one in my head before I'd have an opinion on how much effect it would have.