I like the idea STA except that you are now doubling the maintenance cost of a single installation. Apparently that is what is killing the T-VASIS in the first place. It all comes down to money it seems.
Another idea might be to have one side of the runway PAPI, and the other side T-VASIS. Then we'd all be happy (half joking). Also a prohibitively expensive system.
Another variation would be to get rid of the white lights altogether and have only red lights with perhaps a couple of marker lights to the side of the red lights to indicate where the red lights would be if you can't see any (because you are too high and would have four whites in the current PAPI design. eg.
O is a white light (xenon white!)
o is a red light.
/ \ is the runway.
OO / \ = very steep
OO o / \
OO oo / \ = on slope
OO ooo / \
OO oooo / \ = very shallow
The point here is that you don't have to differentiate between red and a soft orangey white. The red light is either on or off (or transiting between the two which is also obvious and useful) unlike the nebulous white to red transition.
Put it another way, if a pilot is making a circling approach and is rolling out on to final, or making a blackhole runway-aligned approach, and that pilot looks up to see how the picture is looking outside, how much quicker would counting the red lights be done rather than looking at them for a period of time to allow your eyes and brain to see then differentiate the colour appearance of the four individual white/red lights?
I think I might be on to something here!
PS My diagrams aren't being spaced properly. When I use more than one space between characters, this BB reduces the space back to one. But I think you get the idea.