Hi Slippage,
The approximation using 2x's period is accurate for H2, and makes the "together" calculation more accurate since the same stair steps (h1) are used. So that is a good way to do it, the results will be as good or better than separate H1/H2 aggregation periods.
I like the simple together and'ing of the last 4 bars, but I think that each of the bars is going to be the same value, since on the H1 stair step, so may be possible to just check latest together bar.
Using this period multiplication vs HTF aggregation period approach can be used for all timeframes, which would eliminate the repainting, everything would be calculated bar by bar in the current timeframe. Staircase can be added using quantization if it is useful. This is just a suggestion as it provides a "continuous" HTF line vs the stairstep. The advantage of this is that the zero cross will be more timely/definitive than the HTF repainting bar, but will be more dynamic.
For the scan I'm still using the HTF for the main moxie line (cyan). And then the doubled lengths to approximate the second line. You can see the lower line in the non-HTF version crosses zero significantly later. This looks potentially unusable. Maybe useful in its own way but definitely doesn't match Moxie. I'll do some testing later to see which seems like a better signal with a single line on D and H timeframes since those are more critical to the strategy. Also to see how it affects divergence.
About the last 4 bars thing, it's the last 4 scan periods. So 4 hours the way I posted it which is probably too many. I mentioned that to the folks I shared it with in the Mastery room, that we should experiment with how many periods we require them together so it doesn't take too long to show up in scans.