While a re-balancing is likely in order, one issue making it worse at present is the problems with ground support fighters - I've got a bug report post ready on this; the tl;dr is that ground support fighters don't get protection from STO units even when they're over the system body.
By design, dug-in STO units are very hard to dig out with orbital fire, unless it's indiscriminate nuking - this entire dynamic is kind of essential, as it's the justification for ground forces existing at all in most sci-fi settings. And the harsh scaling of collateral damage, and resulting importance of infantry in limiting it is not only "realistic" but a common sci-fi trope as well; witness the famous comment on "killing only left-handed redheads" in the book Starship Troopers.
The problem is that ground support fighters - or aerospace fighters, if you prefer - are a key element in most any sci-fi orbital invasion for these exact reasons. Once they hit atmo and start tangling with similarly-sized ground units, they're within a hulking STO unit's minimum range, and thus are perfect for hunting down STO units and destroying them - basically a SEAD/DEAD mission. (In fact, we might ask for an additional ground support fighter order, in the vein of "Search and Destroy" and "Flak Suppression" orders; one devoted to "STO Suppression. ") Without the ability to prepare a landing zone with a carrier crammed full of tiny but effective ground attack fighters, your only options for dealing with defensive STOs are indiscriminate nuking (which wipes out the populace) or just rushing with your troopships and hoping for the best. (Technically sustained NBG with long-range energy ships is possible, but not on any planet with a decent defensive terrain modifier. )
The "rush" tactic makes all this more problematic, as it means you're either hurling lots of small fast troopships at the target to saturate defenses and get enough through, or going the other direction with a few very heavily shielded and armored ships to ensure they can penetrate. To take a planet without hideous casualties mandates maximal use of infantry with light weapons, given how collateral damage scales - the former approach is sacrificing some of those troops, and building small, fast ships is not very compatible with infantry formations, which are pretty bulky. And the latter approach just begs one to use heavy armor - while it's twice as expensive, it's half the transport size.
And that brings us to the actual details of ground support itself. One might reasonably expect a capital-scale energy or kinetic weapon on an orbiting starship to make more of a mess than one might like, even with careful co-ordination from ground-based forward observers; but a dedicated ground support fighter should work. . . pretty much like real life ones do, under similar JTAC guidance. And if one accounts for the need to limit heavy firepower when attacking highly populated settlements, the ground forces will be highly reliant on precision air support to deliver heavy anti-tank (AP) firepower, which their own formations will lack.
In sum, while the collateral damage model could likely use some simple scaling adjustments or perhaps more sophistication or weighing, it's also true that we don't have a full picture of the at-launch ground combat model of Aurora C# yet. Most significantly, NPRs using their own air support isn't implemented yet, which is significant, as one would expect the NPRs to also have a vested interest in limiting collateral damage on their own planet, and for the reasons discussed above, air power would be an ideal way to do that. As for re-balancing, heavy AP weapons should probably be treated gentler for collateral damage purposes; they're designed for maximum armor penetration on a point target. It doesn't make as much sense for them to cause indiscriminate damage in the way bombardment weapons would. This would resolve the big extant problem of needing to bring effective AP firepower on heavy units, but all options for doing so having significant limitations and/or collateral damage issues.