Given how one-dimensional thruster engineering is, I'm strongly in favour of this. Normal thrusters scale speed, acceleration and handling equally (same multiplier applied to same mass/optimal mass factor), with only Enhanced deviating a bit (speed scales more strongly). I think handling also equally applies to both pitch/yaw/roll (turning) and lateral thrusters.
The only advantage of how one-dimensional this is is that individual ships fly in distinctly different ways that you can't do much about. But overall I think it'd be more interesting if you could choose to dampen a ship's weaknesses or lean in to its strengths.
Personally I'd love it if there were more different types of core thrusters than just Thrusters and Enhanced Performance Thrusters – I'd dig it if we got deep into vectored thrust vs vernier thrusters vs control moment gyroscopes.
But more realistically, they should balance Clean Drives (they currently lose too much performance and draw more power; I think they should draw less power and be competitive on handling, only losing out on speed/accel against Dirty), balance Drive Distributors against Drag Drives (Distributors should be noticeably better on a heavy loadout), add a new engineering effect (call it Gimballed Drives) that improves handling more than speed or acceleration, and an engineering experimental (Supercruise-Optimised) that improves handling and heat in supercruise specifically.
Then you should get a wide spectrum of viable builds – e.g. Clean/Supercruise-Optimised on an exploration-focused Asp X, Gimballed/Drag on a pirate Corsair, Dirty/Drive Distributor on a combat-focused hull tank FAS, Gimballed/Supercruise-Optimised on a Type-9 freighter, etc.