So, Tom, you just make small changes to the bubble count, observe , and repeat?
No, the bubble count is rather imprecise.
A nice vernier handle(these can be DIY with a little paint and add a fat knob etc or buy one etc........) is a better tool for this ona good quality needle valve.
The only other way I know is a mass flow controller, these run about 1200$, but are extremely accurate(should be for the cost).
The rate of CO2 is independent of effects on pH/KH in the tank itself, if you use pH.........this can and does move around, as can KH.
However, the absolute addition of CO2 or the ppm is relatively stable with the valve method. Only degas rates(so keep the tank topped off good) that might change, circulation(you do not want clogged filters etc) or increase/decrease in plant biomass and uptake become the main players(keep tank pruned consistently).
So basic maintenance is more an issue using this method as with any, but not much else once set.