The glide angle of a stopped engine is steeper than the idling engine. Is anyone really suprised at that fact?
It's not quite the no-brainer you imply.
A feathered prop most definitely has less drag than an idling engine, otherwise we wouldn't set "zero thrust" at a power above idle when simulating engine failure on a twin. If you imagine starting at a stopped, feathered prop and moving it gradually to finer pitch, you may reach an angle at which the stopped prop has the same drag as the one attached to an idling engine. Whether you reach this point before or after the pitch reaches the fixed-pitch of the real prop is debatable.
Hartman has a useful summary table for a prop he tested in 1934. With the combination that he chose, the prop attached to an engine throttled to idle was less draggy than the stopped prop at 25, 50 and 75 mph, but was more draggy at 100 mph, presumably because at that speed the prop was effectively driving the engine, while the stopped prop was just sitting there.