Comment 7 for bug 1439928

I don't see any reason for this to be a "Medium" bug - there's absolutely no negative impact described here. In fact, the documented behavior is as-designed. To quote myself from the code review above:

  The most obvious conflicting use case I can think of is having two services of the same type in the same region with different guarantees around performance, availability, etc. Perhaps those two services are branded differently (different service names), and incur different billing behaviors.

Furthermore, if there actually is a user experience issue here, a deployer could solve it through endpoint filtering (exposing one service to one subset of users, and exposing the second service to another subset of users). Otherwise, any bug that could result from this is either because the user is ambiguously choosing endpoints and / or the client isn't providing sufficient feedback to cater to the scenario.