…and I don’t mean the quick wins. Nor questionable substances.
Why, it’s any given pet’s ❱❱❱ speed, or initiative.
It’s always been so much all or nothing.
Being either too slow, or too fast at any given scenario, is wasted points.
And the equal speed is a recipe for disaster.
(Coincidentally, a few surgical nerfs to pets have been applied through moving the bias from Power to Speed.)
Of course, a fight rarely is just 1 vs 1, less so an opponent being known, so a slighty more encompassing remark could be: “A chance of being faster when an opponent enters in”… which does not change the highly circumstantial nature at all.
We like to see fast pets with speed-defined abilities as pleasantly (or oppresively, if extreme) built. But enter something faster than that, and your little wonder becomes a wet noodle. Similarly, slow pets often end up being butts of jokes but you can’t deny in such a scenario they have nothing to lose.
Okay… feeling like repeating the same thing over and over… so do the metas… figures.
So the two main notes to consider are: 1) speed comes at expense of other features and vice versa, 2) speed creep, like any creep, is dangerous for the general reception of everything “below”. Could also add 3) speed is critical for so many things.
What would be the possible solutions to these issues?
To address 1), I’ve thought of giving speed points some actual value, such as an explicit “bonus for each N points above/below the opponent”. New cases of creep, flyers and even Surge/Powerball users would be a giant red flag instantly, so it’d be capped at, like, 20-30 points. It would also give more incentive to maintain speed shift teams which often feel “not too serious”.
To address 2), there is always “squish everything”, and at this point, as I consider a bigger deal to be made, this would be par for the course. But other than that, and more important to provide a failsafe for the creep in the future, there needs to be a better compensation for being slow (And it isn’t gluing a crazy high burst damage to everything.). Leaning into 3) as well there are a lot of abilities which are better while faster on the top of natural initiative mechanics, while the opposite cases are scarce and almost never a focal point because avoiding to reveal one’s intention is this much important.
What would be a decent compensation, though, or another solution, I don’t have many ideas, and will blatantly rely on you folks for the best of everyone. 
It is also tempting to question the mechanics itself, but at this point, it would change so much it’s time for a 2.0… and I am not opposed to that as long as it’s a clearly defined event… Hmmm.