This is what is nice about research. Real problem solving matters.
But also what is frustrating about research: real problem solving is hard!
Another set of games my girlfriend recently turned me on to are rhythm games, particularly VOEZ[1]. Again, as I don't think I said in the first comment but said elsewhere, it's not like I don't like the "classical music" style and I'm all "jazz music," but my "jazz" style is definitely my default mode. If you asked me afterwards to remember the notes, I would be able to tell you the gist but I certainly don't remember details, a lot of it is focusing when it gets complicated and responding on beat. And, as I said above, my absolute favorite is playing a song for the first time and trying to get a good score, because I really have no idea what's coming and it pushes my "respond quickly" buttons.
[0] not that you don't learn quickly/during the fight without dying, and as many games often are, there is an obvious hint of what to hit, the rhythm to a boss, etc.
I totally relate to that guilty feeling. I like having to think on my feet, and never saw the point in memorizing a sequence of controls. It's like it ceases to be a game at that point!
It's just that some games have this strange curve in the learning curve, for instance Starcraft, where beginners think the game is all about strategy. But as you learn to play, you find out that it's more about macro, cause having more units normally beats any strategy. But then if both players have top notch macro, they have to begin doing tradeoffs in what they build, maybe cut some units to get a technology faster, or be greedier but more vulnerable for a while for a later payoff. In other words, the game becomes about strategy.
wikipedia.org/wiki/Mornington_Crescent_(game)
My completion of Age of Engineering is this small countryside feeling open air base, Direwolf20 (famous YouTube Minecraft player) beat it with a space station. The only obvious commonalty is we both have huge floating energy spheres from Draconic Evolution which looks kinda out of place next to a greenhouse compared to inside a space station. "There's more than one way to do it" is a driving ethos in most of these packs.
For speedrunners though any sources of randomness are mostly just annoying. If it's 50 seconds faster to start on a hill, and your personal best is 40 seconds slower than the world record, you're not going to play runs where you didn't start on a hill. You're just going to reload.
I remember feeling guilty when I remembered what was on the next screen, because it felt like cheating. It kind of broke my heart when I realized that was the best way to get good and it was pointlessly stubborn for me to resist a technique that games were designed to cater to.