Yes, that is the detail behind what I meant by it "complicates things". I was referring to those things complicating my otherwise simplified focus on a comparison of kinetic storage to battery storage, but didn't want to get into the specifics in that post. The momentum-to-momentum storage/spending and the added speed mean that it isn't just a simple "which form of energy makes the best storage" discussion.
But, having gone there now with details, I'd point out that these same considerations are also an argument for *applying* energy on the downhill. In other words, if these effects are important and a net benefit for added speed then zero electrical output may still not be the answer there, potentially positive energy output could be warranted. And, on flats before a hill, potentially increasing energy output might help...? Has anyone every tried that or found it to be true? (Searching online for bicycle advice on climbing hills they don't recommend increasing speed before a hill, but I'll admit that I've done that because it feels better to my novice muscles. Perhaps I'd feel differently if I was as highly trained as a tour cyclist?)
But, does any of this fall on the side of gaining speed (and inviting more air resistance)? I'd like to see math and/or experiments before I take it as a foregone conclusion, but it is a more complicated bit of math than simply looking at steady state downhills or long flat sections.
And how much speed will you gain and how quickly will it dissipate? If the hill is long enough to reach a steady state then you lost some energy you could have saved with regen. Do we then calculate how much speed is optimal and how far from the end of the downhill to end the regen so we can gain exactly that much speed? And for what % of the total drive time on a given trip does that calculation matter?
In the end, though, while we are discussing if and how often coasting is a useful tool, I'd still point out that regen is often an optimal answer as well and so eliminating regen (OP point) isn't the best control configuration either. And if coasting is a good answer a lot of the time, is there a more natural way to achieve it by looking ahead and using good observation and situational awareness and rules of thumb about what you see on the road than there is by having the accelerator end at the coasting state? Would having "foot off pedal == coasting" provide the driver with a crutch that will make them want to use that tool more often than it is the best answer?
At least you admit there is a lot of information that we don't have. I agree those are all very good questions, and for lack of information, I haven't argued about whether regen is better than plowing into the wind while speed doesn't increase further. There has been a lot of talk about "you could be using regen instead of fighting wind" and that argument may be accurate, but it seems counter-intuitive to me. I don't understand how one could gain from regen while also rolling at a speed where wind-resistance prevents further speed gain, because gravity is a constant, so either regen will reduce speed (because it is additional resistance against the same gravitational constant) or wind won't prevent speed gain (allowing regen to absorb the remaining gravitational pull, but you could gain speed instead). However, I don't have data, and it is certainly complicated, so I'm not arguing.
Speaking of data, my very first post on these forums was a link to a video of the energy chart doing something that doesn't make sense to me, and it was never responded to. What the energy chart is doing may well be accurate, but I believe that what I capture in the video implies something is incorrect that may or may not be covered in these three possibilities:
1) The "average" line is labeled wrong (it isn't an average, but some other function).
2) The "average" line and historical graph come from different time samples (possibly of the same length, but not covering the same start and end points).
3) The vertical axis for the chart is not consistent with vertical axis for the average (the numbers or units of measurement actually differ even though one set isn't displayed).
It is also possible that the chart is wrong, in which case the data could be wrong elsewhere as well, and that might even make certain testing moot.
Now, regarding the pedal off equals coasting discussion. The OP request (regen disabled) is not an option I personally would expect to ever use, but that doesn't mean it shouldn't exist (I say for user preference, but technically, it might be justifiable even if solely because even low regen could be dangerous in some weather).