> Closest approach time [...] t_max = 2023/08/15 12:22 TDB [...] Run started at 2023-08-14 23:51 UTC and ended at 2023-08-15 00:06 UTC

So this computation result became known after the closest approach would happen at the latest.

Looking at the asteroid I saw hitting us, 2023CX1 (known as Sar2667 before being designated), it says:

> Closest approach [...] t_max = 2023/02/13 03:22 TDB [...] Run started at 2023-02-13 11:47 UTC and ended at 2023-02-13 11:53 UTC

Again, hours after the impact actually occurred. And that's from the 7-observations page, there are 3 more pages with 28, 76, and 125 observations, all listed here: https://newton.spacedys.com/neodys2/NEOScan/index_past_imp.h...

The next one on the past impactors page has a 3 day delay between when it hit us and when its orbit was computed

What's the point of this?