I'm working on an RTS too, there isn't too much to show yet.
The "inspiration" is supreme commander and other RTS being low on the complexity and planning aspects.
The approach I have seen in the industry is that people take AoE2 and starcraft as a baseline and then switch out or improve different elements. E.g. starcraft 2 massively improved unit movement and pathing. I think total biscuit tried to make a mod where resource gathering was "automated" and easier and more recently "battle aces" focuses more heavily on the skirmish aspect. Many opinions I have also heard boil down to "if you remove micro, you remove the game". And that's not wrong, I can certainly see the point and the skill differentiation between someone who can and someone who can't micro their units.
But what I want to see is all of things that people already do "in their mind", like picking a build order, certain defined "points" in their own "gameplan" that they decide "X units A Y units B is when I should attack", or "transition points" and steps, and to make all of that explicit.
MTG "deckbuilding" works the same way, players anticipate certain problems and situations and then they build their decks with specific setups in mind and situations that they want to reach, and if they reach those states, victory gets very very close.
Taking all of that into account, surely there are just "strategies" that work better than others and finding those is more interesting or at least equally interesting as micro to me, but basically no games give you any tools or help to actually do it. You basically have to take out pen and paper to write down what worked in your last game, what didn't.
What would a game look like that gave you ALL the planning tools and performance metrics?
To me, that's where the modern "big scale" RTS fail, or rather, why they don't interest me.
And also, once things are "perfectly" planned and prepared, there are always ways to introduce e.g. random failure into plan steps to keep players solving "micro" problems, they would just happen in a different place.