Short answer: It does not say anything!
Longer answer: Both the number of submitted papers and the number of accepted papers don't mean anything independently. There are topic-specific conferences and their communities are relatively small. Also, there are researchers -for a reason or another- who submit according to the location where the conference is held. All these reasons make it normal that some conferences don't receive a lot of submissions.
However, the ratio of the number of accepted papers to the number of submitted papers can be an indicator of the conference quality. It is actually used by a lot of conferences as a quality advertisement.
In the end, I wouldn't compare the conference 'C' against conference 'B' because this comparison is most of the time subjective. I would rather submit to a conference that has a good reputation and is well ranked (e.g. core) or has a good h-index (e.g. google scholar).