Even if a miracle happened and we reached zero emissions today, the climate will keep warming at least for some more decades due to climate lag and feedbacks.
See this comment I wrote about the permafrost:
This is based on estimated total biosphere emissions, estimated human emissions, and annual CO2 PPM increases. I did some math on that and wrote it up here if anyone wants to dig deeper: https://biotinker.dev/posts/climate1.html
In the case of the zero-emission miracle, atmospheric CO2 levels would start dropping at about the same rate they are currently climbing. As you mention, there is climate lag and feedbacks, but the "return to normal" would start fairly quickly.
For example, there are only rough estimates on how much methane is trapped in the permafrost. The first estimate I saw by Natalia Shakhova et al was 1400gt over a decade ago (so about 45 years of human emissions at the current rate). I found this recent MIT article which mentions 1500gt [1]. Let's not forget methane has a 25x more heat trapping capacity than CO2. Do the math on that. Also the permafrost is just one of many feedbacks.
Another point you're not considering is that the capacity of the biosphere to absorb carbon is being reduced. For example, the Amazon used to be considered a carbon sink but it is becoming a carbon source [2].
I'm not convinced on your second point; I'll need to see more data. I read the paper you linked, it doesn't give hard numbers, so it's difficult to compare the net loss of carbon flux in the amazon, to gain of carbon flux in other areas that are getting net-reforested, such as the eastern United States. Examples of local areas flipping from sink to source, and others becoming larger sinks, don't paint a clear picture of the directionality of the whole system.
The most recent IPCC report (AR6) gives the "very likely" (90% probability) range for "equilibrium climate sensitivity" - the eventual effect on temperature of doubling CO2 - as between 2C and 5C. Current CO2 levels are less than double pre-industrial levels, so the "baked in" warming could well be less than 2C. Also note that reaching the equilibrium temperature takes more like centuries than decades, making the concept somewhat academic, since lots of other things will change over that time period.
Also note that 2C warming over pre-industrial temperatures would just take us back to about what the temperature was 8000 years ago, when conditions were good enough for agriculture and civilization to start.
Have you ever read about the Russian wildfires of 2010 which had a massive impact to global food prices which then triggered the Arab spring revolutions? That was a single event in a world below 400ppm but these heat waves are starting to occur with more frequency now and will have serious consequences.
The IPCC is an organization that functions with consensus and their data is awfully outdated. Their assessments generally underreport the consequences of climate change. Here's a paper from 2010 which analyzed this: