My first reaction is: I think the simplest hypothesis is a continuous tree ring record. It's continuous everywhere else. A sudden gap needs more than a just-so-story about Romans to justify it.

Also:

There are enough samples to satisfy professionals that they actually have the whole 2000 years covered, but the sample overlaps for the gap between the blocks are few and rather weak.

That sounds very much like fitting the evidence to the hypothesis.