A recent playtest with Ingenero resulted in an opening battle between Nazis and MI6 agents in the Pariasian catacombs, mostly over trying to rescue a human sacrifice.

Anyway, the battle itself just seemed to go on and on, but not in an interesting way.  It was just a battle of attrition type of deal. 

The rules for Ingenero are specifically set up to discourage battles of attrition and encourage use of fictional positioning and risk taking.  They do this by making only a minimal level of harm inflicted for straight up attack/defend interactions.  To increase that you must take risks (basically upping the ante in order to deliver more harm).  And you can make risk taking less risky by first getting your character into a position of fictional advantage. 

The reason why this combat degenerated to a drawn out battle of attrition is that either side did those things.  Part of that was the players were new to the rules and didnt go for any advantage plays or risks, and part because I (GM) didnt play the NPCs that way either.

I hadnt played such a large scale battle before with the rules, and it was hard to keep track, as the GM, of multiple character with multiple objectives - as in different characters going for different advantages and what was this or that other guy doing again? and so on, so I just defaulted to retaliating against whoever attacked who.

In retrospect, I think the way to play it as GM when you have lots of NPCs (whether they be individuals or groups of mooks), is to think about just one objective for your 'side' as a whole, and have every significant character or group of mooks support that one objective at a time.  Then you have just one problem to think about rather than trying to keep track of multiple ones simultaneously. 

I seem to remember reading through an Ingenero summary you posted here a couple of weeks back, and it seems like it's supposed to be quite cinematic and action-packed. I don't remember how position and advantages worked, but from what I gather from your post, players (/NPCs) basically choose if they want to go for causing damage or vying for advantage which doesn't cause damage now but will pay off in the form of more damage later. In any case, I think you're on to something with thinking about the NPC combatants in terms of what their side's goal is. If my understanding of advantage is correct, what might make it easier to control groups of hostile NPCs as a GM, could be to group their advantages together. To take a concrete example: Two mooks lay down covering fire towards the player characters so that two other mooks can move into a better firing position.
This may be more of a general GM tool, but I find that running combat in this way - especially post-WW1 combat - can be quite effective in getting the player characters to react to what the NPCs are doing, instead of coming up with a tactic amongst themselves and sticking to it - or indeed acting completely independently. What I mean by this, concretely, is that NPCs communicate. Loudly. Firearms make lots of noise. If you want your squadmates to hear that you're throwing a grenade or that first squad is supporting second squad, the enemy is probably going to hear it as well (especially if you're fighting in close quarters). If the NPCs aggressively vie for advantages and the PCs hear "Smith! One frag grenade! Fire!" "Frag out!", they're going to be doing stuff left and right, and it's probably going to be more involved than simply saying "I attack."

Yep, agreed totally - reactive NPCs suck.  Next time I will make sure that the NPCs lead by example and execute their own plan to achieve their own tacical goals rather than be the gun-wielding equivalent of martial arts mooks who mill around waiting for their turn to take the hero on 1:1