It would be a thrill to see a continuo player of the 18th century prepare his part, and to ask him, for example, why he over-figured one bar and omitted all the figures in the next one. The logic of original figured basses does not often match modern expectations. But in many of the scores I prepared years ago, I can actually find the same kind of logical lapses. We tend not to notate the things that seem obvious to us, and we over-notate anything that scares us. Perhaps continuo figure frequency, whether historical or modern, is something like a sophisticated personality test.
In any case, in a score like the Bruhns cantata I just looked through, around 20% of the continuo figures are missing, which is not a very great problem. I usually like a complete set of figures in my part and I like to identify and correct possible original mistakes, so I take my time and work my way through the score, pencil in hand.
The part that really boggles the mind is the attitude of the average modern editor. Here is a person who’s being paid for working out a continuo realization (in the Bruhns, this was in fact rather neatly done, apart from a curious persistence of translating 5b into a seventh chord instead of a 5-6 chord), so the editor is in fact working with: the score, the bass line, and the figures. This edition faithfully reproduces the original figures with all their original omissions, until on page 13 all of a sudden an editorial 6 in brackets appears. After this point, the editor adds perhaps ten or fifteen more figures here and there, but not nearly all of the missing ones. Why?
To be sure, this riddle is not a huge deal in terms of the usefulness of the score, but for anyone interested in the workings (or non-workings) of the human mind, it is extremely curious. Did the editor work his way backwards through the score but lost his spirit on page 13? Are there some hidden rules for the apparent randomness in adding figures, such as “you have twelve free pages. After that, add a missing figure every 7th bar, skip four bars forward if there’s no figure missing in that bar, skip three bars backward if there are two figures missing, and so on.” I can’t get it worked out.