Supporters of a published topical position statement (camp statement) can indicate support by attaching their profile name to the position as supporters.
The system appears to have been designed with one support vote for one position, per topic in mind. Each vote by an individual supporter is diluted correspondingly for each additional support vote he adds for alternate camp positions in one topic.
While this may well serve scenarios in which such makes sense (topics where each camp position is actually different and competes with all other camp positions on the same level), it does not make sense for surveys in which an individual may equally support various camp positions at the same level (sibling camps), and believes they should all be included as supportable parts of the indicated topic.
Surveys whose object is precisely that inquiry, (how many, of a total population of support for a particular party's platform, support the inclusion of each or which of the planks identified?), are valuable research tools that may well be useful to many potential system clients.
Examples of this type of survey research include this suggestion box, and the Platform of the People (a generic political platform where each "camp" is actually a platform plank)
In such cases it does not make sense to dilute each support vote for camps by the number of camps supported by one supporter. Nor does it make sense for the sum of all individual votes for specifically indicated camps (planks) to be represented as the total "votes" for the topic (Platform of the People), because more than one individual camp support vote may be made by one individual.
It is useful to know the whole story and represent the whole story in a way that is easy to understand. For example, of XXX number of people who support the Platform of the People (tally of individual supporters of the platform, without duplicate counts) in general, XX of them support the inclusion of an Abortion plank (tally of individual voters who specifically support the inclusion of the abortion plank).
This type of survey is a quasi multiple choice survey where not only one, but many of the multiple choices are expected by each supporter.
While architectural support for this particular type of research survey may not be the original or primary objective of the Canonizer, it is most certainly a survey service that is widely needed and has the potential of attracting users to the Canonizer.
Note: This suggestion overlaps somewhat, but not entirely, with the Modified Tally Count suggestion