[tialaramex]

But it isn't trivial, as I explained the leap days inserted to fix the calendar are entirely predictable. In 1752 when my country adopted this system they already knew that 1800 and 1900 would not have such a day but 2000 would. They had no idea what the year 2000 would be like but their calendar already told them it has an extra day.

In contrast leap seconds are not predictable because they're trying to smooth out spin variation and the planet's spin is affected by natural processes. The IERS measures UT1 and we add (or remove) leap seconds to make UTC try to approximate UT1, we get a few months warning at most that a new insertion or removal is needed.

And despite being very difficult all of this is also pointless because we don't care. As an astronomer you know you don't try to use a wristwatch to figure out the Earth's rotation, you can just look that up online and use your exact position. But for any other purpose except astronomy we don't need this angle information, we want a nice regular time and the atomic time, TAI, already provides that without this nonsense about leap seconds.