Incidentally, the Unicode conversion error you identified for 0310 is spot on, Greg. I remember having had issues with converting the (old) word files inherited from Michael Welte into the database for the first incarnation of the NT.VMR. Upon realizing the problem we had already manipulated and updated the data such that a new clean import was considered to be the less desirable option. Hence, I went through all the entries where Greek letters were to be found and manually corrected them. This entry was missed, probably because it looked kind of ok.

Unicode conversion error should be added to the arsenal of mechanical errors in the transmission of texts, in case it has not been listed already. The terminus postquem of such errors is pretty easy to narrow down for later generations, I suppose. Probably easier than the errors resulting from transitioning from majuscule to minuscule script.