The nice thing about celcius and kelvin is that they're the same scale, but celcius is just shifted 273.15 units. And it's more intuitive for humans to work with smaller numbers with bigger relative differences. But yes, kelvin would be a lot better to work with, especially considering stuff like doubling temperature (doubling energy) would actually work correctly in kelvin.

But if there's one thing that makes a lot of sense to base temperature enough for human use, I would indeed say it's water, because all life uses water, we are completely surrounded by it, and it's super important to nearly everything we do too.

Sure, but the vast majority of people live in low lying areas and even then it doesn't shift that drastically. You need to climb a mountain to see the difference when it comes to applications of daily life.

Although now that I think about it. The same criticism applies to pretty much every definition of temperature that is based on the behaviour of matter. This also applies to Kelvin. Temperature is a property of matter and every type of matter behaves differently.

Technically Kelvin is an absolute scale, where there are not "degrees," just references to "absolute zero" which is supposed to be no vibrational energy, but is not actually zero kinetic energy since there is still zero-point energy. It's as arbitrarily chosen a reference as any other scale, and for a specific purpose. We used it often when I was in the natural sciences, but it can be just as strange to have to think in terms of adjusting for 273.15K for the misleading "freezing point of water" or 298.15K for STP, another arbitrary standard of measurement. Kelvin is no better than Rankine. And, it's even more confusing if you think of temperature as an average energy in a system, and have to consider quantum gases with a temperature "below absolute zero."

Annoyance in the 21st century at cultural relics and colloquial is kind of silly.