I wouldn't be surprised if they didn't notice how broken they are because they're plain bad at the game. The unfair spawns is one thing. But that can be seen as random (although they obviously suck at testing if AZ is right, and he probably is), and something you race to get too.
They don't seem to see items as MUCH better in Brawl. Items in Melee sucked. I'd rather chuck them off the edge than use them most of the time. Or chuck them at someone. Using them was suicide. This is not the case with Brawl, as every item has gotten a buff to the level of broken. But because they aren't throwing out constant hitboxes, they've never seen an item literally spawn in someone's hand. Playing with their ruleset against someone, an item spawned in one of our hands (usually the one being chased and run down) 3 times. I cancel out my tumbling animation with an attack. Due to the low hitstun, this allows me to use Pit's nair pretty soon after getting hit. I was picking up an item as it spawned.
Occasionally, I would be using Pit's nair offensively, and as I was hitting my opponent, I'd get a free green shell in my hand because it literally spawned there. They may call this stage control. I consider that unfair. I had no forseen way to know that item would be in my hand.
Then I had one match where the entire match degraded into chucking one home-run bat at each other for the entire 3 minutes. The item never disappeared, and was involved in every kill. The home-run bat is incredibly powerful when thrown, and was thus a better kill move than any of Pit's attacks.
Oh, and people who argue about the time (Yuna), stop. Time's not an issue with the items on. The matches go a lot faster. Think about what would happen if you gave Pit a projectile that could kill at 80%. This projectile can be angled to go below the stage. The time is only an issue if they turn items off mostly. Alternatively, it may not have taken long because my Pit rushes the opponent a lot and uses an arrow when he can't catch the person.
EDIT: Very strangely, I just did the same test on FD. Two Marios. First test is one Mario suicided twice. The spawns SEEMED about the same (I didn't count this one though, but that's because I kept getting a 1:1 ratio when I was counting.
Then I had one Mario kill the other one twice, and separated them to other sides of the stage. The winning Mario either had a starman spawn on his side, or if a starman landed in the center, I counted whichever side it moved towards (aka whoever had the best chance of getting it). It was 10:3 in favor of the winner at the end. The video was 2:57 when I finished, so it's recorded.