(a) time estimate are dangerous. In one case the leader of our team estimated 5 people for 5 months. The company wrote a fixed-cost contract. After 10 people for 18 months the company folded.
(b) time estimates are meaningless. All of the software I worked on was new. There was no "prior art" to ground the estimate. Projects were almost always "late", mostly due to changing requirements.
(c) time estimates assume a known future. Who expected the pandemic? (Or the Spanish Inquisition?). How good is your estimate when the 6 people you could "hire" are all rejects from other projects because they can't program?
(d) time estimates assume a known good tool set. CorelDraw won't run on your old Mac but the company won't pay for new hardware. Microsoft pushed out a "destruction tuesday" update and now your machine won't boot. Somebody "upgraded" the library software and now your code won't compile. And don't get me started about chasing a compiler bug (I've fought 5), each costing an unplanned week. Or hardware bugs (3 times), that basically stopped the project until they were fixed.
(e) time estimates will cost you time and money. Suddenly everything become a crisis. Cancel that vacation. Work nights. Work weekends. Nobody gets a raise if this is late.
(f) time estimates WILL be used against you. Managers write down the "estimate", stick it in a drawer, and when "blame time" comes you can be certain to be "to blame" for a bad estimate.
I know you think I'm wrong. You "just know" my "lived experience" is nonsense.
You can't estimate software. Get over it.
Read "The Mythical Man Month".