This is an interesting paper (and thread). I could have been expected to think it correct because it validates the models (well, one model), but there are a couple of issues that give me pause... https://twitter.com/hausfath/status/1354504732291731456

The new reconstruction is not based on any new data, but it calculates an adjustment to the existing ocean SST proxies (mostly Mg/Ca or alkenones) by examining how much those proxies change in a previous period - the Last Inter-Glacial (LIG) (around 125,000 years ago).

Those changes are assumed to be due to variations in the seasonal insolation (which was larger & more rapid at the LIG than in the Holocene). If the changes are most coherent w/one season than another, that is then applied to the Holocene records to get an estimated annual mean.

This method assumes some kind of linearity of response to seasonal insolation across the two interglacial periods, but doesn't show that the lifecycle/seasonal preference of forams involved would actually provide this response.

I will be looking for some forward modeling of the behaviour of these forams to validate this statistical technique. (Full disclosure, this is something I worked on ages ago, and I'd love to catch up with what's been done since. https://pubs.giss.nasa.gov/abs/sc09000a.html )