Having done quite a lot of chemistry and having a pH meter that reads to .001 with a temperature probe....
I would suggest that quite a lot of the bouncing is not only the good suggestion of plant waste but also the sheer volume of the solution at hand and the minute changes you're making.
Even with a run to waste system, I find the tank pH will change quite a bit over a few days. Particularly near the start of a newly mixed tank.
When you add the pH down (acid) you are altering the ionic properties of the salts in the nutrient mix. It takes a long, long, long time for all of it to mix and homogenize 100%.
What tends to happen with mine is I'll start with water at 6.8. Adding some nutrient might bring it down a few .1's to somewhere closer to 6. I then add acid to bring it down further to the 5.7 / 5.8 mark.
Over the next 24h+, it'll gradually go back towards 6.
When the acid first goes in, you can imagine it like dropping food dye into a bath, even if you're swirling it with a pump or airstone. It might mix well, but it needs to mix to a molecular level before it will finally sit still. The pH bounces up as the acid becomes 100% homogenized and changes the ionic makeup of the solution.
I'm interested in the idea of really carefully setting your pH to optimize growth, but the level of information available for specifically tuning cannabis is low from what I've seen so far. And anyway, I doubt it really has to be that precise, I'm sure there are a lot of other areas that could be optimized first and produce a greater result for less effort.
If pH bounce is pissing you off, a buffered solution might be something to look at.
In the same way that you can make you own pH up and down solutions, you can also make your own buffers. Buying it will probably be a far easier option for most. The chemicals aren't expensive or difficult, but it requires some chemistry to work out the amounts needed for a specific solution and pH.
Dissolving (even if it's a liquid into a liquid) is something you can't trust your eyes and senses with when it comes to fine detail - like honing an engine out and saying "yea... that seems close enough" will leave a lot of room for it being shit. When I was doing higher level chemistry, we'd often make up solutions of things and then test our skill at working at concentrations of other solutions by using one as a standard. To make this standard, we'd precisely measure out a known amount and concentration of one chemical and then dissolve it in water or alcohol. Even working with 100ml volumes, we'd have to shake and turn the flasks for a few minutes at least to make sure there weren't patches of uneven concentrations in it.
With tens of liters, tiny pH changes and more things going on than I can count in the solution, it's natural that it can take a lot of time.
You can expect bounces to last considerably longer if you add anything as a solid, even if it looks dissolved. Your concentrated nutrient mix probably looks liquid to you, but mine tends to produce some very fine silt if I leave it to stand for a while. Those same tiny specks will need to dissolve over time before their affect on the pH will become entirely clear. If you like my reply, pls vote in my bulb poll (sig).