Earlier quoted context omitted.
Nowhere in the "Expected climate impacts" section is inducing an ice age even on the table. The anticipated effect is a 2C cooling of the atmosphere. Considering we're on track to blow way past a 2C increase in atmospheric temperature from 100 years ago to 30 years from now, SAI would be a mitigating action. The goal is to maintain our habitat.
"Expected" - is the keyword here. No one should (and I belive will) agree for any kind of "geo-climate-engineering experiments" unless our science is actually capable of predicting/simulating exactly what's going to happen. Said understanding of the climate and ability can be proven by let's say consistently successful weather forecasts for 14 days in advance. Then we can talk. Until then, even seriously considering…
Weather prediction and understanding the impacts of SAI are two very completely different problems. This is an arbitrary and not particularly relevant benchmark for gating geo-engineering work.
That being said there are many, many ethical, political and other technical reasons why geo-engineering is a very bad idea and should be avoided at all costs. Undoubtedly, mitigating as much as climate change from occurring by quickly reducing global GHG emissions should be one of the main laser-focused areas of everyone involved in stopping climate change.