From an engineering perspective, the 'best' long term approaches are those which produce either the largest number or largest percentage of desired outcomes. Also, getting people to change their historical mindset is a big part of effecting progress. To me, if you get people treating their pets and food animals better, 'performing' or 'display' animals would follow course.

Singling out one species with a tiny number of members in a very unusual and rare use case may be great for splashy awareness campaigns, but unless whoever you're making more aware is also likely to carry that over into their treatment of other animals, I feel like you've fundamentally failed to begin to address the biggest part of the root problem.