In fact, here [1] is a well sourced article which highlights at least one indication that this is the case - historic CO2 estimates derived from plant stomata are much higher and closer to contemporary measured levels than ice core data. Then there is the idea of time averaging smoothing out peaks, and we can only speculate/model the loss of resolution.
It's very easy for a well intentioned group of researchers to reason themselves into a corner, particularly when things like institutional momentum, political biases, and understandable doomsday concerns influence the kinds of questions scientists ask and the paths they are willing to take to find answers. Because this research has broad effects on economic, social, and corporate policy, it is dangerous to close off one side of questioning behind an apparent consensus, not to mention it's simply bad science.
>the time resolution of the ancient ice core data is quite low, while all the anthropogenic action happened in the last 150 y or so
All the more reason to be open to questioning.
1. https://wattsupwiththat.com/2010/12/26/co2-ice-cores-vs-plan...