There's not really an alternative measure by which to measure the quality overall of the games as a whole that EA the publisher is backing. Until a better measure comes along, Metacritic is the least bad available option. I'm not going to try to argue that Metacritic is the be-all end-all when it's not. However, currently Metacritic is the best means available to obtain a cross-section of critical opinion about a given video game.
Why shouldn't multi-platform games count? Each platform version of the game is a different SKU, deployed on a separate piece of hardware to optimize for. Sure, it's the same game high-level, but making the same game run well on different hardware takes unique time and effort. I know this first-hand.
Outside of the one Kane & Lynch game which got a Gamespot journalist fired, I'm not sure how much of a leg the implication of this point - that video game reviews are paid off for good publicity - has to stand of. Games for big-money IPs can earn poor reviews - the most recent being Aliens: Colonial Marines, and also Tron, almost every 3D Sonic The Hedgehog game, and Duke Nukem Forever - and new IP games can get great reviews - Portal, Braid, Limbo, Bayonetta, Super Meat Boy, etc. One would expect well-known IPs to consistently score the highest if the publishers with the money always scored highest. Further, well-known game critics such as Adam Sessler wouldn't be as visible and respected as they are if there were any evidence of their being paid off.
This can be made into a fair argument, but be careful to be consistent. By this same token, it is unfair to criticize EA the publisher for publishing any EA Sports title; the publisher just backs the title, the developer makes it. In the case of Madden, EA Sports publishes the title, Tiburon and EA Austin develop it.
A counterargument, however, would be that once Bioware is purchased by EA, they are in fact EA and should be included in the EA envelope.