There is also example (10)f, with two candidates marked and one with an additional mark through it, somewhat like the disputed ballot, but not exactly (ballot format in the example is different).
The context of other votes on that ballot is also ambiguous. Other partisan rates appear to be all R candidates, but one of them shows a filled in bubble with an additional X through it, similar to the filled in bubble with the / through it for the D candidate in the recounted race (the R candidate in the recounted race has a filled in bubble).
I think it’s too ambiguous because of the X through the single candidate in the governor’s race – so I don’t think it should have been counted. But apparently is a done deal at this point, and there are probably other ambigous ballots in the recount that were accepted, or not, based on judgment calls that could have gone the other way. (Remember hanging chads? And the Al Franken recount?).
I can actually see the logic behind the random draw at this point … you could probably have another recount with a different set of election judges and have it come out another way. At least the drawing resuls in a clear winner that cannot be challenged, though not exactly democratic. (But either way, it’s pretty much a 50/50 tie between the candidates— so whoever wins will have won the support of half the voters in their district).