Advice on implementing highly mixed spots

Splitting between multiple cbet sizes is too difficullt and unnecessary for microstakes. We don't lose much EV by simplifying our flop strategy to one sizing so that's usually ok.

To find out if we should we can measure EV lines. You set the solver to take one action exclusively (e.g. cbet 33) and then compare that total EV with the standard solve. If the difference is within 1-2% of the size of the pot it's too negliable to care and you can use the simpler strategy.

As for remembering lines, patterns > specifics so even if you get the combo wrong if you were thinking on the right lines that's usually a good sign. You want to hit the flop and be able to say, roughly, what you'd do with OPs, what you'd do with TP, what you'd do with air etc. Obviously you can't predict turns and rivers that will drastically alter plans but that's the art of it, changing lanes mid hand because you had such a clear idea what your range looks like. Somewhat easier in 3b pots so it's a good place to practice.

You're tying to build some replicable narrative in your mind so, for example, when I hit this flop I'm thinking, first, can my range cbet? T-high I imagine we can probably polarise a bit and have some larger sizings. UTG will have a hard time calling all their broadway overcards and our range benefits from folding out that equity. I would assume the EV of betting either b75 or b50 is neglible and I would tend towards using b50 as a default for convenience, knowing it won't affect overall EV much. This frees me up on the turn too, because when he calls i have a better internal picture of my range.

i also know in general it's better to slighly overcbet this spot because it usually is in cbet opportunities, as populations play poorly vs them in general (or more likely misplay their calling range) so i would tend to go slightly linear and bluff anything with future equity. These are simple heuristics but I know they aren't going to be a massive mistake (which is what mixed hands are ultimately, not that we're too worried about getting every combo bang on) and i have range clarity in my mind so that when he raises it's trivial to play back at them.

Ceres put it well. If the equilibrium strategy doesn't bet closer to everything and there is quite a bit of checking, look at the property of the combos that like to bet often, or those that don't bet much at all. Do they have overcards with 1 or 2 backdoor draws? What about hands with draws + showdown value? You don't have to care about being close to balanced, just create patterns to find the hands that solver likes to bet the most and bet them pure if you don't have good reads on your opponents or population. Then compare the SBvUTG ranges to SBvBTN ranges and see if there is a difference and why that is. Focus on the things that makes you go "hey, I've studied this, OOP needs to do this and IP needs to do this. Does IP actually do this? Is it natural? Can I exploit that?". If you are confused about sizings and overall betting frequencies, look at range equities and who proportionally has more of nutty hands etc. You will find patterns.

If its a mix in theory that means EVs of mixed actions are the same in theory. In that case you ask your self, how should my opponent play so the one or the other option becomes the best.

If you are trying to balance your own range then one size is the best approach, complications of using multiple sizings are overstated esp in lower stakes. 99% of regs dont pay attention on what are you doing, so you dont lose EV even if you are unbalanced.