I never really understood why we need leap seconds. Or better: why we need to bother with them in a computer system at lower level. If we decide that we absolutely need to keep our time in sync with the rotation of the earth, really what should be done is define a timezone with all the leap seconds applied, and use that timezone to only display it to the end user. Not change the way we sync computer clocks for no rea…
The cost/benefit at the time didn't look so bad because the world wasn't full of distributed synchronized computer systems, so the added 'cost' of leap seconds was just some make work for geeks in national timing labs.
The cost benefit is very different today.
We can keep civil time roughly aligned with the sun by moving timezones an hour every four to five thousand years.
Applications that want to give accurate sidereal time or very accurate sun-up sun-down can use predictions of UT1. Bonus: it's a lot easier to give an accurate UT1 when you don't have to worry that leapsecond (mis)handling has screwed up your underlying clock.
That's another part of the cost model that has changed: When leap seconds were created it would have been burdensome to carry around an additional offset in time transmissions for those few applications that want a more accurate mean solar time. But today its fairly easy.