If your goal is to not have a mess, you can get there by cleaning up all your messes. You can't get there by not adding to the existing mess.

There are messes that cannot be cleaned up. We have the leverage to destroy everything. Therefore it is better to focus on preventing the mess, than in futilely pretending we can clean it.

Obviously it doesn't make sense to clean up messes that can't be cleaned. There are also messes that can be cleaned, and it can make sense to do so.

It's no substitution for actual data, but there is a basic logic to it. If you're bleeding now it's more important to stanch the wound than it is to worry about avoiding future wounds (they're also not mutually exclusive conditions).

Not sure analogies are strictly helpful here. Like, alternatively: Say there's a knife being pulled through your guts. Should you prioritise stopping the bleeding or trying to arrest the progress of the knife? The analysis here needs to be more quantitative: what are the returns on various mitigation and cleanup methods, and what do the respective rates of change imply in terms of comparative environmental impact over a fixed period of time?

50 > years of pollution that will maybe take millenia to clean up by themselves?