Exactly, it's the production of the phones where most of the carbon is generated.
As for the comparisons there, it's hard to make head or tails of the meaning, because the "context" provided is apples to oranges. A million cars versus how many phones? Or a million cars versus how many cars in Europe? Citing a big number does not mean much unless we have something to compare that big number to, and I question the intentions of such "bald big numbers" because I see it so often used to muddy the waters and confuse people on many topics. For example, nuclear advocates will often disparage solar, talking about X million square kilometers without giving any context on whether that's actually a big number or not.
2.1Mt CO2 (over how many years with that "by 2030") versus how many Gt Co2 per year? Seems to be 2.5-3.1Gt/year based on a web search..
And what percentage of total EU emissions are from cars?
These numbers also make the argument that phones are a small rounding error compared to other much bigger actions.
Should we mandate repairability? Of course! It's good! But getting bent out of shape on the impact of phones compared to far more common and wasteful practices is a bad use of our limited amount of time to make drastic action.
And the EU is far better on cars than the US, so perhaps it makes more sense for them to take action on phones, but in the US, our car addiction makes for far better and practical environmental action.