Hi @nleahy ,
Firstly thanks for the question and your patience.
I've looked at the App-V doc and according to this architecture overview app-v publishing servers are designed to work in the way you are describing. In that you can deploy publishing servers on multiple computers for disaster recovery. That being said my understanding of the docs is that the correct way to do this is to replicate a publishing server and then use a load balancer to direct traffic instead of on the client specifying a list of publishing servers. What I'm saying is that according to the docs your use case is supported but in reverse, the redundancy should occur with the publishing servers and a load balancer and the client configuration should focus on one publishing server. That is the architectures intent.
----------
Please don’t forget to "Accept the answer” and “up-vote” wherever the information provided helps you, this can be beneficial to other community members.