.... I measure that somewhat by the amount of twist required to maintain a speed. ...
I don't think that's a very useful measuring technique.
If we made the assumption that we needed the same power whether in 4th or 5th to move the bike at a particular speed, the engine is obviously turning faster in 4th. So it needs less air/fuel drawn into the cylinder for each stroke because there are more strokes per mile; in the lower gear, you'd need the throttle set lower to get the same total air/fuel per mile, so your test becomes dubious.
That also immediately says that engine pumping losses would be higher in the lower gear due to having more vacuum in the inlet manifold - one of the reasons diesel engines are more efficient is that they work at full throttle all the time, so run with minimum pumping losses.
Where the higher gear gets less efficient is when valve timing becomes unsuitable and/or the air and fuel aren't mixing properly for complete combustion at very low speeds.
I've occasionally tried using my bike's mpg figure to compare 4th and 5th at a speed held steady on the cruise control, and a few attempts at odd speeds between 30 and 70 give me the impression 5th is better than 4th, but short term variations due to slope and wind make this difficult to get hard results.
Oh, yes, pumping losses are mostly what gives you engine braking when you shut your throttle. If you get the opportunity, get up to speed, shut your throttle, get the feel for your engine braking. Now try again, this time turn the ignition off, open your throttle fully, feel the engine braking. I think you'll find it's significantly less. (Won't work for a Gen 3 - the throttle plates won't open with the ignition off. Damned YCC-T.) Oh yes, if you do this test, shut the throttle BEFORE turning the ignition back on.
But you already know you get more engine braking in 4th than you do in 5th.