Min-maxing player characters in adventure games where that's a viable strategy tends to be unpopular. I suspect the reason for this is that the burden for making encounters challenging to a group of min-maxed PCs falls entirely on the referee. Whereas players have to only min-max one entity (their characters) to get the full benefits, referees have to repeatedly min-max entire groups of enemies, and may even have to create several such groups for a single session. In systems with leveled power progressions, players also have time to incrementally develop their knowledge of how to build and play the character, whereas the referee is usually going in without the chance to playtest an encounter beforehand.
Therefore, I would like to suggest the following non-exhaustive list of strategies for handling min-maxing in games.
Copy Characters
Hold onto character sheets from old campaigns (or ones who have died in the campaign) and reuse them, but this time as NPCs or even reskinned as monsters. Since the PCs usually don't see the stats for monsters, they won't often spot that a monster is actually a reskinned former PC, though they might notice that it uses a similar set of tricks to one.This usually isn't a problem, but rather leads to interesting problem-solving, since the players know how the power interactions work, they can work to neutralise them by interrupting the interaction. You can even split up a character sheet for a suitably high-level PC and create several monsters out of it, especially if they have more than one combination of powers they use frequently.
Most groups in my experience have one person who likes to play lots of different characters, and is constantly drawn to new rules material when it comes out, and tries to create some version of whatever they think the hottest, newest thing is. It can be extremely useful to channel this person's interest in trying out the material by having them build new potential characters, even if they don't want to play them right this second. You can either use the old characters, or the new characters, whichever ones they're not using at the moment. By reskinning them as new NPCs or new monsters, you can do this without "giving away" which particular set of stats and powers is being used.
Monster Colosseum
Every three or four months in one group I was in years ago, we would skip our regular session and instead play "monster colosseum". Everyone got a CR budget (this was in the D&D 3.5 days) and picked monsters out of the Monster Manual until they reached that budget. Occasionally, to mix things up, we would roll on the premade encounter tables organised by CR in the back. Two people (of four or five, depending on who showed up) would draw the map on whiteboard, we would each place our monster teams, and then we'd have a giant battle to see who's monster team would be the last standing.
This had the effect of basically show-casing monsters (especially weird ones we might not otherwise use), and highlighting how to use them most effectively and imaginatively. e.g. one guy once got a bunch of giant wasps and used teamwork between them to pick up the enemy monsters and drop them to do tons more damage than the wasp sting alone. The referee gets to see how different monsters and different combinations of monsters work with other people doing the bulk of the work figuring that out.
Keep Records
If you have an encounter that works particularly well or particularly poorly, it's worth writing a short note to yourself about it after the game, and storing these notes over time. e.g. "Skeleton archers on hill that's tough to climb, too easy b/c of flight powers" or "Giant wasps can't injure PCs b/c they can only be hit by magical weapons" is about as long as they need to be. If you play using player roles, you can assign this to whoever keeps the notes for the party, and have them turn the notes over at the end of the session. When you're next planning an encounter, you know what to change or preserve right away, rather than having to puzzle out what went wrong based on faulty memories of previous sessions.
Similarly, referees should recycle rules material they know well. For example, I recommend drawing up a couple of spell lists at the start of a campaign and reusing them frequently. You can describe the spells differently each time to change things up, and you can gradually focus on increasing the number of spells you understand extremely well, but having a few spells you know like the back of your hand means less page-flipping, a smaller burden on your memory, and more time spent figuring out how to use the spells in imaginative and interesting ways. I recommend something similar with monsters - learn a few monsters and their powers extremely well, reuse them extensively, and then make minor variations on them as needed.
Lastly, if you notice something works extremely well against the PCs, make a note and figure out how you can exploit it in different ways. This might be a low saving throw, a lack of a particular ability, dependence on a resource that depletes, etc. Write these out on a sheet with each PC's name and particular weaknesses underneath it. If you have space, you might want to list their strengths as well. That'll give you an idea of what powers to look for when you're putting together encounters. Do the PCs have crappy willpower stats? Then maybe mind control monsters that attack those stats would be effective. Rather than having to evaluate a ton of rules material and generate off-the-cuff optimal strategies, you can simply hunt through the available material for the bits that seem directly relevant to dealing with those weaknesses and strengths. Especially combined with the method of encounter recording that I mentioned above, you should be able to winnow through rules material much more quickly, or design off-the-cuff powers that directly exploit those weaknesses.
Conclusion
The more work that's done to shift time away from poring through rulebooks and trying to figure out the rules side of things, the more time can be spent making encounters interesting and memorable. Similarly, the more familiar the referee becomes with the rules, the greater the set of affordances they offer to the imagination. For these reasons, I therefore suggest you do whatever possible to lighten the cognitive burden of optimising encounters.
