Yeah, that's the stuff I was talking about forgetting. I'll try to remember, but some forgetting is more-or-less inevitable after several months of playing nothing but Guild Wars 2 and Satisfactory - I might have slightly burnt myself out on SE [and other 4x/wargames] between our game and (mostly) the beta testing.
Remembering to attach them to brigades was one ah-ha moment, as was making sure the brigades were in a posture like Blitzkrieg for Attack and AP bonuses. Not taking them back out between turns even though it made them safer was another - that costs you a lot of readiness. For the first 10-15 turns, I was shuttling around one commander in Blitzkreig to be present at every nuke launch regardless of whether they were on the north or south front, but then I'd detach the nukes during your turns to make them safer and make it less obvious what I was doing b/c you didn't seem to be using similar postures for your artillery. That not only denied them the AP bonus from the posture, but also cost them readiness from detaching and attaching the subunit - and IIRC readiness directly translates to a modifier to your attack value, so the difference b/tw 20% readiness and 80% can easily be "I killed one guy" and "I killed one stack". Besides not swapping in and out of formations, just taking a turn off was often useful for actually being able to hurt my targets. Nukes weren't getting full readiness recovery between turns, but (if I can be forgiven for using made-up but broadly indicative numbers) regening 20% per turn from, say, 10% to 30% each turn meant the unit was never going to be worth a damn but if they took a turn off, gained 20% while using none, and then and then regened 20% from 50% to 70%, they could maintain the same rate of fire while being twice as strong. Again, these numbers are made up but the idea holds - and this really compounded with the aforementioned lost readiness from constant reassignment.
At least one later ah-ha moments was the depressing realization that unlike real heavy artillery, the best kill ratios in SE come from areas where troop deployments are very thin. Concentrated troop deployments protect the troops from getting hurt by spreading out progressive defense decay as non-lethal hits wear down surviving units rather than being spread across the many troops piled up within the blast radius. This is 100% a factor of how SE treats artillery as a series of weaker successive attacks that randomly hits something in the hex each time rather than a strong AoE attack that hits everything there (or a portion of everything there) with a single blast. That arguably makes sense for artillery, up to a point - artillery operates in barrages even if this still makes bigger stacks safer from artillery than small ones. However, nukes are just one really big blast, so this doesn't even work conceptually. This inspired me to stop trying to blast large stacks of troops unless they had something really weakening them, b/c otherwise I'd barely scratch them. The units with compositions I wanted to blow up basically laughed off nukes in my tests, while the units I did vaporize would either be really weak, momentarily vulnerable, or only a couple of troops. It's why the times I managed to sneak a nuke into range of your HQ units they were vaporized - well, that, and the fact that HQ units don't get the leadership bonuses their subordinates do, so their subunits would be like 10x weaker than subunits in the line units I couldn't scratch. You may have noted I had all my command units in huge piles, often with pure garbage militia, but also line units held back from the front - this was why. Soaking up artillery subattacks made the concentrated stack harder to wear down - and keeping the line units there helped avoid cascading failures as readiness/morale/fortification/etc were gradually reduced, subunits died, and the remaining attacks were concentrated on fewer and fewer, weaker and weaker troops.
The last ah-ha I can remember was that nukes in particular aren't great for weakening entrenched units prior to an attack - but they're absolutely devastating in wiping out routed units after a victory, even if the victory basically just pushed the defending stack back w/o killing much of anyone. The loss of fortification combined with morale and readiness loss would be the difference between a nuke killing 1-2 subunits and then the conventional attackers pushing them back with slightly fewer losses taken and slightly more inflicted but most all of the defenders alive to fight another day... and the conventional forces needing to fight harder to push the unit back but the nuke then destroying 80-100% of the retreated troops. The nuke going first makes conventional hits more likely and nuke hits less likely, while reversing them swaps those - but of the two, only nuke hits have an autokill chance. This is also why making a bloody but unsuccessful push with conventional forces followed by nuking the victorious defenders usually worked better than softening the defenders with a nuke before attacking.
Put together, you end up with some really counterintuitive outcomes. Nukes are terrible against massive formations of fresh troops in fixed positions... but they're great at wiping out small, sparsely-deployed formations or large amounts of scattered, mobile troops. The same things are true of artillery & aircraft, but they don't feel quite as ridiculous b/c those are at least multiple munitions over a period of time causing multiple strikes within the AO.
I'm pretty sure I've forgotten at least one other tricky bit gleaned from testing & in-game experience, and verified by subsequent usage. If it randomly comes back to me, I'll pass it along, but I'm not optimistic.