Part of the beauty of live performance is the risk of messing up! It just takes practice to mitigate mistakes as much as possible (it’s never perfect which is also fine) and to learn to accept what we cannot control. But that’s not what you’re asking.
What you’re describing reminds me a lot of live “dub techno” sets, which are reallllyyy engaging (to me) and are essentially a live mix on stage. The tracks are precomposed and stems/individual tracks are usually played back from a device (SP404, Octatrack, Laptop) but the performer/producer will use effects and jam out with all kinds of things - EQ, other samples, remixing on the fly, etc.
Part of the beauty of using electronics is that we can control the degrees of freedom that we have when we’re performing — the more freedom I have in my mixing during performance, the more freedom I have to deviate from the predicatability/reproduced accuracy of my music. In other words, I have the freedom to play with things if I want to, or I have the safety of using my predetermined composition without risking mistakes or deviations.
But like others have said, it matters less to me what someone is doing on stage than what music they’re making. Performance has value, and can also manifest in myriad forms. One of my favorite performances was a guy walking on stage, plugging in his iPod, pressing play, and walking off — it was conceptual of course, but it felt unifying, and the music became the focus instead of the person on stage with flashing lights and stuff.
Another idea that I’ve had a lot of success with is playing a stereo track through multiple channels (surround sound) and manipulating where the sounds go in real time. It can be really beautiful if it’s done well, and while the musical content is the same every time, the audience’s experience is still moment-to-moment. Hard to perform in most spaces, but if you can get your hands on a close-to-identical set of 4-8 speakers, I highly recommend it.