Epoxies seem to share some characteristics with concrete as far as curing
Exactly, and for much the same reason.
Concrete's curing depends upon water and carbon dioxide molecules finding their way to the uncombined calcium oxide to form hydrated calcium carbonate. Once the mixture solidifies, the only mechanism for this reaction to occur is for previously combined molecules to uncombine, passing the CO2 and H2O along and recombining with fresh CO2 and H2O, in effect passing the molecules along like a bucket brigade. This process will continue as long as there are uncombined sites but will necessarily slow down as the number of uncombined sites is available.
The University of Wisconsin began a study of concrete over a hundred years ago. Every year, a sample was removed and tested for compressive strength. The study was to have ended after fifty years but their testing showed that the samples were still increasing in strength so they elected to carry the study for another fifty years. As I recall, the strength continued to increase throughout that part with the study ended because they ran out of samples.
The epoxy reactants are an epoxide and an amine. The amine has multiple reaction site on each molecule which permits the cross linking that gives epoxy its rigidity. Like concrete, in a liquid form the molecules are free to move around to find active sites but once solidified, the bucket brigade mechanism is the process by which unreacted molecules find partners. And like concrete, as fewer and fewer active sites are available, the curing rate becomes slower and slower.
Fortunately, with both concrete and epoxy, the products reach an acceptable state of usability long before the reaction is complete. In my particular case with the epoxy composite, we were trying to make a product with extremely tight density tolerances so curing shrinkage was of great concern. I would think that shrinkage, hardness, and possibly temperature resistance would be issues of concern when trying to build a precision surface such as the floor of a milling vise. To that end, I would error in the direction of overcuring rather than under-curing.