My expectation "ab initio" has been that E.G. is not truly multi tasking. That is, it will not initiate multiple synchronous tasks on the server. Indeed, I can think of circumstances where this would be less than ideal, such as when the output of one process defined the input of another. It is reasonable to expect that in this case, you could enqueue multiple tasks, but they would still execute serially.
The huge benefit of E.G. over Foundation SAS is that you are given back your project to work on while your asynchronous task(s) are still incomplete. So, at 4pm, when you have a large analysis to do, with three different ways to do it, you can enqueue the three different ways and return in the morning to find the 3 sets of results and choose the correct set of results.
I run jobs locally, and on one or more servers depending on where I am. Because I can change servers, I can actually initiate three concurrent processes if I send them to different servers, but I think we are pushing the limits a little there, especially since a number of clients are exchanging local installations of Foundation SAS for an Enterprise installation of SAS to be shared from a big server.
Needless to say, running three different processes is only something I do rarely.
Kind regards
David