@OwenD Just got verification that this works (at least to sense the filament not running out).
... I increased it to 100ms to be pessimistic, but that would be correct... the printer is still (for a little while) in filament. It should run out for real (testing the other branch) before this print is through.