Okay... are we sitting down comfortably, because this could be a long one... (long sip of coffee)...here we go...
What I mean is this: In your program you could compile together a LIST of entity names that you want to hold together and instead of using that LIST to form a GROUP, you could just keep that list separate.
Why would you want to do this???
Well, think of it this way. If you have it in a GROUP you are stuck with some of GROUPs more limiting features. For instance, if you want to change the layer of only one of the entities. Using a selectable group, you end up changing all of them.
With using a stored LIST instead, you could conceivably do anything you wanted to the individual entities (move, trim, copy, extend, mirror, etc...) and then using a command that makes use of the LIST you could then manipulate all of those entities in anyway. Of course, to do that you would have to build commands that would explicitly deal with that list.
Then again, on second thought what would probably be the most useful to you instead of a list is a global SELECTION SET. It's used the same way, but then you wouldn't have to test to see if an entity from your list still exists before trying to perform an operation on it.
Does this clear anything up any?? Or is it still clear as mud?