The hard part about trying to add slope-like rules is their directionality. There are only a few places if any at all on a hill where you can see to lower levels through 360°. In most places you can see smoothly down the front of the hill but not back behind you, which either stays level or rises up to a peak. And in most places you can't even see sidewards to lower levels. If you have mostly one-hex-wide contours, level rules work well. When you want to have wide contours but depict a more even slope instead of a wedding cake, you need something else. One way to get the directionality is to mark hexsides, as slopes do. Any other solution must, in my opinion, have some way of determining directionality of the permitted LOSes. If a solution does not have directionality, it fails what I consider the prime function that slopes perform and cannot be considered a suitable replacement for slopes.
Another prominent feature of slopes is that they can see OVER half-level obstacles such as walls, hedges, rubble, etc. This is a secondary consideration, but for any slope-replacement rule would have to handle this if it is to be used as a replacement in existing scenarios.
The TEM for slope pockets is a really small part of their functionality. I think it could even be removed from the existing slope rules without affecting very much at all.
With the wedding cake model, units only have LOS to lower elevations when they are at a crest line. Slopes increase the number of hexes where units have LOS to lower elevations, but in fact unless a map is really marked up with slopes, they don't really simulate a smooth hill much better than crest lines without them. But using slopes is a bit better, at the expense of complexity.
One might also consider increasing the number of hill crest lines by decreasing their modeled height, but then you would have to start adjusting the heights of other obstacles by counting them as 2-3 slope-replacement levels for each standard level. You would also have to adjust the number of blind hexes per slope-replacement level in a similar way.
If a slope-replacement rule passes the first two tests in a simpler way than slopes, I would take a good look at it. The "I can look in any direction from any hex as if at the peak of the hill SSR" fails the first test so badly I couldn't see it replacing slopes. It also fails the second test, but the first is far more important.
JR