I would have started with the data sets.
The reason why is that when you change a rule. That rule will effect every datasheet in the game. This can lead to really unbalanced interactions (like the one pointed out). Or strange exploits, like a guardsmen calming complete immunity to a bloodthirster, by virtue of standing on a ledge 1 inch tall, that the later can not fit it's base on.
Your issue with changing sheets is there's probably over 1,000 sheets overall, each with stratagems and fluff you need to represent while trying to maintain balance and considering how effective that cost is.
Here's an example of what I'm talking about.
Slaanesh, Nurgle, Tzeenth and Khorne are in my codex. If you make one too strong then the codex gains redundant choices. If you keep them all the same then the fluff isn't represented. If they're strong in combination, if you design them to be used this way. Mono-god becomes redundant, if you give bonuses for being mono-god the, undivided and possible 3 of the mono-factions become redundant. - And so on and so forth, this is just with one book.
If you were to say make krack missiles flat 6 damage, then then all multi-would models become less effective and would need a points reduction or defensive buffs. Each solution creates and snowball of implications.
Here's another LoS. True line of sigh punishes things that are just too big to hide, so do you consider the size of the model when generating it's stats. Or use generic rules where the size of a model is redundant, meaning wraithknights are leaving 1mm of there base in terrain borders to claim a save.
The TL:DR is it's absolutely massive task to balance 40k
Edited by Battle Brother Abderus, 28 September 2021 - 02:26 PM.