Yes, I agree that a tiny difference in efficiency isn't going to be worth it if it costs important apm. I have no idea how popular an idea it is to buff higher tier engies to reduce t1 spam, but I definitely dislike the huge swarms of t1 engies, so I would prefer some sort of change that discourages that. We might see some positive changes simply from t2 and t3 engie buffs. If t2 engies are a bit more efficient, maybe you will send half of the 30 (or whatever) t1 engies assisting your hq that is upgrading to t3, up to the front on a suicide reclaim mission and replace that bp with some t2 engies you build out of a t2 support factory. So you could end up with fewer total units and better pathfinding. Maybe it will still always be better to just use factories at the front to get the reclaim asap, though. Or maybe people would just slightly adjust how they increase their build power, and making an extra t2 support factory or two to make bp later on ends up being slightly more optimal than making t1 engies a bit earlier, I'm not sure.
In any case, if it is still not worth the apm to adjust your bp around, that would just mean that any buff to t2 and t3 engies just needs to be a bit larger to have any effect. I think the problem is all tiers are EQUALLY mass efficient (10.4 mass per buildpower for all engies), so there is zero benefit to building the higher tier engies FOR BP (so you only make a small number to access higher tech). The only exception I can think of is where you really need to concentrate tons of bp in a small location, perhaps to assist an smd or shields, where a huge mass of t1 will not all get close enough to assist.
Maybe changing the cost of t2 engies to 10.0 mass per build power, and 9.5 mass per build power for t3 would be a small change that makes them more viable, without making them too much better either. So t1 engies would be still 52 mass for 5 bp, t2 engies 125 mass for 12.5 bp (reduced from 130 mass), and t3 engies could be 285 mass for 30 bp (reduced from 312). That makes t2 engies less than 4% cheaper, and t3 engies less than 9% cheaper, so those are pretty small changes we could try very easily.