I’m curious what your use case is here. The individuals servers, as with most scaling load balancing solutions, should be able to handle any request, and in fact what you describe means they would be able to do that very thing. Allowing hangfire to handle the balancing means you don’t have to go back and configure things when you add servers 3, 4, and 5. You also wouldn’t run into the case where server 1 is doing all the work while server 2 is idle just because all of the jobs currently queued are 1-10.
There are times when maybe you want a specific server to always be the one to get a certain request without backup in the case of something server specific like maybe you only have 1 box that has certain physical files on the machine and if you want to do that, you can accomplish that by creating a queue for that job that only the one server has defined in its options.