I'd have to load up a game to get an exact number. If I can be bothered. But it is a stupidly high amount, with a "perfect" import and all DLC.
Dug up my old spreadsheet. Our theoretical EMS maximum was 4303, all DLC included. (I wouldn't say this has been extensively vetted, though.) Not quite 7K, but 1200 points over what's needed for all endings. A playthrough with no import and all DLC gives 3601 points, so that's 500 spare points. Note that maximum w/o DLC or an import is only 3004; IIRC that means Shepard cannot survive.
Is being able to skip content with no penalty a reward or a problem? Depands on the game. But to me, a game that supposedly celebrates a choice-filled narrative that gives the player the ability to shape the story, getting the exact same outcome whether you actually followed the story or played a few dozen hours as a faceless N7 sounds like a problem.
More realistically, the effect would be to permit players to skip the scanning and fetch quests. Sidequests aren't all that productive in WA terms anyway, so I don't think making further WA awards superfluous would change the incentives to play them. But I suppose we should ask someone who actually plays MP what effect, if any, it has on his playstyle.,
Though personally the main problem is: no matter what you do, you are only allowed to do so well. You can't hope to aspire to an ending that sucks less than what we got, no matter how "perfect" a game you play, no matter how much DLC you get, no matter how big an arbitrary number you get. You simply don't deserve better. No matter how big of a difference you made, it doesn't make any difference.
I don't see how this has much to do with EMS itself. Any other system for determining what ending you get would also only give the endings that the designers decided to give you. Unless you're saying that being able to exceed the existing thresholds gave people false hope? I agree with that, at least to the extent of dynamically adjusting the thresholds to increase with DLC. Not sure about how to handle imports since I can see a case for letting importers have an easier time of it, but the counterargument is pretty strong too.
(Please don't tell me you fell for that Puzzle Theory nonsense jUliA used to push.)