Sure there are the energy losses you cite - air resistance, rolling resistance, and various mechanical losses such as hub friction. But some of those are independent of weight and none of them depend on whether the weight is located in the rims or in the rest of the bike & rider. So these losses have to be compensated for by our pedaling to exactly the same amount whether we have a bike with a heavy rim and light frame vs. the same total weight by with a light rim and heavy frame.

Yes, there will always be slight accelerations/decelerations as we're riding, but those don't involve extra energy losses from heavier rims. I.e. the bike with heavier wheels will decelerate a bit less if you relax your pedaling for a moment due to a 'flywheel effect' - the bike with heavy wheels will coast a bit better and keep a more consistent speed. As a result, even though a bike with heavy wheels is harder to accelerate, you don't have to accelerate it as much and the result is that the energy requirement to propel a bike with an extra pound in the rim is the same as a bike with the extra pound in the frame except for the brief initial acceleration (and subsequent accelerations after braking).

And sure, weight matters when climbing as I indicated in my previous note - but it matters essentially the same whether the weight is in the rims or in the frame/water bottle/belly/etc. No matter where the weight is located you still have to get it to the top of the hill. The only difference with weight in the rim is the tiny bit of extra energy when you first get it up to speed in the first few seconds of the climb since that energy (stored in the spinning rim) is lost when you put on the brakes after getting to the summit. You can get a good idea of how little extra energy is needed by putting the bike in a stand and using your hand to spin the cranks and get the wheel spinning - takes only a second or so of easy arm effort to get it up to speed compared to the hour or so of both legs working to get you and your bike up a decent climb.