> but also would almost certainly cause environmental impacts at least on the scale of climate change, and possibly worse.

Or, possibly better. The deep oceans don't have a lot of marine life because there's not much plankton, because all the nutrients have sunk to the bottom. Regions where there are natural upwellings that bring bottom nutrients to the surface tend to have massive plankton populations, leading to massive fish populations (Grand Banks, the west coast of South America, others).

That's why it would be important to do experiments to see what happens when it begins to get scaled up.

"We don't know what would happen" is not a reason to avoid experimentation (indeed, "we don't know what would happen" is pretty much the reason why you do experiments).