I have been following advice I saw to only charge when you are below 100 miles. I am charging to 80%. So now most cycles are about 30% to 80%. The extrapolated charge is back to 308 miles with 247 miles of charge at 80%. I charged one time to 100% @ 307 miles. It took a few cycles to have it correct. The car has almost 7K miles.
Mine has 6,600 miles currently, and charging to 90% every day for over a week has been giving me an extrapolated 100% charge level of around 295 lately (down from around 300 that I was getting initially when charging to 90%). That's better than the 288-ish I was getting when I was charging to 75% or so regularly, but still a little bit too bad that it's below 300 miles. Ambient temperature that might affect capacity hasn't been an issue, either, around here in the SF Bay Area, where we've had an uncommonly warm fall in the parts I inhabit (as in 75 - 85 Degree highs virtually every single day for a long time).
Last thing I'll try, is the let-it-go-down-to-30% plan you describe above, charging only when the car gets down to 100 miles, and then fill it to 80%. The only thing I don't like about this approach is that it goes against Tesla's advice to keep the car plugged in every day, which you can't do, at least not on the days you're getting close to that 30% level (Since the lowest I can set the charge limit to, is 50%). Still, there might be a happy medium if I set the charge limit to 50% (which should result in a plugged in car but no charging on days it's actually above 50%), and then set it to not charge at all once it goes below that level, until 30% is hit (then charge to 80%).
In the end, I just want to do this sort of test (before it gets cold) just to verify I've got appropriate capacity, after accounting for some amount of actual degradation. One reading in the 305 range would be nice before we head in to the "cold" months of 40-ish degree weather . . . . we'll see.