This is all jolly interesting stuff, bringing up questions of how to model modern tactical C2 on the tabletop that are always worth thinking about.
I was going to suggest permitting companies to activate as many times as the owing player likes in a turn, but imposing stress points only on those platoons that take actions. On the supererogatory embranglement side, this means the additional recod-keeping burden of recordng stress points for each platoon, rather than each company. However I think it should offer the compensatory advanatges of solving your “platoons spread all over the map” problem, and also giving a way of keeping a reserve at company level.
The idea of operating in company-sized sub-units is a good one, although I wouldn’t insist on their consisting of exactly three elements; whereas Western armies task organize by cross-attachment (an infantry company swaps one of its platoons for a troop from an armoured squadron, for example), Warpac armies always used to do so by augmentation (for example break up a tank company and attach a platoon of tanks to each infantry company of a battalion).
You mention the problem of support-arm and recce elements, and it is true that it makes more tactical sense for these to be able to act independently in platoon-sized elements than it does for maneouvre arms. One method would be to activate them along with the company team they are attached to, which makes sense for sappers, anti-tank, air defence and artillery tactical parties. It does not make much sense for recce, though, who by the nature of their job ought to be operating well in advance of the main forces, and maybe in even smaller than platoon elements. I have watched a brigade-level exsercise where the brigadier tracked the position of his battle groups by four-figure grids, one for the BG centre of gravity, and another for the location of the BG recce — and for the British army, the BG recce platoon/troop would be just 8 cars, operating in pairs. One possibility would be to allow recce to magically self-activate. Another woud be to represent higher levels of command as pseudo-company-teams, and activate elements reporting directly to that level of command as a company group. For a British orbat, that would mean activating the recce platoon/troop as a BG-level asset. This method would fit quite well with the Warsaw Pact habit of giving commanders at every level above battalion some kind of recce element under their direct control.
I know you want to avoid representing higher HQs, and indeed I think it would be quite possible to drive yourself potty trying to account for all the HQs present n a modern brigade, where each unit might have a main, a tactical and a step-up HQ. However I’m thinking that it might be possible to represent some aspects of electronic warfare by permitting EW elements to apply stress points directly to HQs, without them having the benefit of an activation.
All the best,