It is expected for model experiment description papers at GMD to show some evidence that the experiment described works as expected. Usually this involves including some evidence that the experiment, or a very similar experiment, has been performed in a similar complexity model with an indication of the level of model performance. Sometimes an experiment with a simpler model may suffice. Have I missed it - I can't see that any model runs at all been performed for the PDG? If there are actually some results, please cite them, and briefly discuss their performance. If not, perhaps we can consider that last DG model experiment results are a sufficient proof of concept? In either case I would like to see a subsection somewhere that summarises the key differences between the two experiments in terms of forcing and boundary conditions. You already have a paragraph in the Introduction that highlights the observed differences between the two time intervals, but that is not quite the same thing... Also, would it be possible to include some indication of model output in Figure 1 (or elsewhere) for either deglaciation?

Please also update the code and data availability section with the correct unique identifiers for the data sets.

Apart from these two hopefully small things the manuscript looks ready to be published.

Thanks for adding the "proof of concept" information - I think it is very helpful context.

Sorry it has taken so long, but happy to accept the manuscript now!