Now I know that fuel senders and gas gauges from this era didn't have a whole heck of a lot to compensate for the contents of the tank sloshing around but I'm not sure what I did this time to make it even worse.
The last two times I had to dig into the cluster for unrelated reasons, the gas gauge decided to misbehave once it was all back together. The first time it seemed "too tight" and for the first tank of gas it was very jerky in movement but it did eventually settle down and become pretty linear with it's wobbling as you turned corners. This time around even though I was careful not to touch the needle while the cluster was disassembled and during cleaning it now acts "too loose". Even getting into the car is enough to send it on a wobbly trip up and down the dial and makes it a lot more annoying as you try and guess how much is left in the tank.
Suspecting I damaged a wire (or one of the riveted terminals, since those don't seem to like being handled excessively), I pulled the cluster again (I sure am getting good at this!), tested the grounds, tested the battery and ignition power and tested that all the wiring back to the sender was good using a spare cluster. The wiring was fine and the gauge itself is actually good. If you test it on the bench an open or shorted signal path pegs it beyond the E of F but with the sender attached it tracks cleanly between the sender end stops, but with no lag or hesitation, so indeed the gauge has become very, very sensitive.
I know some gauges compensated for extremely sensitive readings by using some sort of a grease that maintained viscosity across a wide temperature range. I can only guess it's been wiped out. Problem is that unlike the Eagle's speedometer and tachometer, the fuel and temperature gauges have the dial riveted on the front and the assembly is riveted on the back to the cluster circuit board, leaving only a small opening to reach the front needle shaft. Having just got off repairing the mistake of oiling the speedometer needle shaft I'm being extra cautious how I approach this problem since if I make a mistake there is no easy way to fix it.
Again, this isn't a sender calibration or an electrical issue. The gauge has simply lost its ability to compensate for variations as the fuel moves around in the tank. Google is absolutely useless for answering a weird problem like this.