The abstract combat system (The Externals, page 64) seems useful and intuitive for a certain level of attention to individual battles within a wider view of a conflict. I'd like to propose a couple of quick fixes.

First reworking Tactics, I could say that a battle of matched forces (0.00 numerical advantage) with competence, but no tactical success on either side should tend to have a similar chance of either equal losses or a greater loss by one or the other side. From equal numbers to a relatively small advantage, a commander should therefore check Tactics at +0.

As one side's advantage becomes greater, it gets easier to force their preferred range and to concentrate attacks by enveloping one flank of the smaller side, though the smaller force could still win on the other flank. The benefit of greater numbers increases to attack both flanks and encircle the enemy, to the greatest possible advantage of surrounding them completely, given hex mapping, at 6:1 odds - but if badly handled, that could just let the smaller force attack in every direction, so there's a limit how much it helps.

Change the Tactical Advantage table:

I want damage to relate sensibly to what's inflicting attacks and to be able to finish a battle. I also want to keep open the possibility of resistance or even success by much smaller forces, so some results should allow a side to get away without severe losses for a number of phases. I propose to do the first just by applying losses by the other side's current combat strength, and the second by varying damage down to as low as 1%, maintaining the pace of combat, which I think is about right, by translating each 5% of damage to a d8%: