Thanks for the response, Haso. Although an upgrade may be the final solution, in the short term it may not be a viable option. Our upgrade process takes a little while 
After some testing in our Pilot environment and delving into the logs, we have stumbled into a specific line of the mrp log that it repeatedly fails at. It is in the portion of MRP Regen that deletes unfirm jobs, and is below:
09:59:48 Site: SITE After Delete of Table: PcValueSet for JobMtl
10:04:38 The underlying provider failed on EnlistTransaction.
10:04:38 The requested operation cannot be completed because the connection has been broken.
10:04:37 The underlying provider failed on EnlistTransaction.
We are wondering if there is a data related issue in that plant (part/routing/job). We were completing successful regens with some regularity until a few weeks to a month ago, so it seems something changed. If anybody has any ideas, we have open ears.
We have a ticket open with Epicor, so if anything comes up from that case we will post here. The help is very much appreciated.