Not really, because most carbon credits are based on counterfactuals - would this forest have been logged if you didn't protect it, would this farmer have used less efficient practices if you hadn't paid them, would this land have remained empty if you hadn't paid to plant trees, etc. No matter how reputable the rating agency is, they still won't be able to answer these questions very accurately.
Additionally, when you purchase a credit you're generally purchasing a prediction of how much carbon might be absorbed in the future. For a afforestation credit, you're often purchasing the amount of carbon that forest will absorb over 100 years. No rating agency can legitimately tell you that your forest isn't going to burn down or be logged in the next century. It's a pretty ridiculous idea - the seller, purchaser, and rating agency will all probably not exist after 100 years, and the government of whatever country the forest is in will probably have been overthrown.
I think contributing money to protect and rewild nature in the interest of increasing carbon sequestration is a great idea that should continue, but nobody should be receiving offsets that they weigh 1-to-1 against their actual real-world emissions. It just isn't a meaningful comparison. Even if your purchase actually makes an impact, it's like dumping toxic chemicals into a river and setting aside money to fund an organization that will gradually clean up the river over the next century, and then claiming that you've now achieved net-zero pollution.
I think you two are saying different things. Agree that the existing products for decarbonization have a lot of room for improvement.
Some solutions include:
1. Stopping the dumping
2. Pay the full cost of the impact of any dumping that does happen. In your analogy, this would mean the longer you leave the river polluted, the more you have to pay to clean it up.
I like this framing. That's because it seems like a working carbon market is a must have to hit decarbonization goals. Therefore it's a question not of "if", but "how". And a lot of the current "how" is probably best priced close to zero.
like can't you have different ratings for different carbon credit products - the more accurate ones are priced higher?