Sorry, Djourou is quality but Koscielny is shockingly bad, they don't work well together, Djourou saves his *** all the time.
That's not true at all:
Arsenal's win percentage with @JohanDjourou starting this season in the Premier League is 79%, without him 40%.
Arsenal's win percentage with Alex Song starting this season in the Premier League is 71%, without him 0%.
Clearly, both of those players are particularly important but Koscielny doesn't need his *** saved by Djourou all the time - Djourou has said before that they're the perfect (modern) partnership, while he brings height and strength, Koscielny has pace and likes to get tight to his man.
They both cover for one another all the time - it works both ways, but Arsenal's win percentage with Sebastian Squillaci starting this season in the Premier League is 53%, without him 67%.
Clearly Squillaci is the weak-link, and doesn't fit into Arsenal's system at all - we play a high line which he can't cope with because he lacks pace, while he isn't good enough in the air to be part of a deep line.
That's why he and Koscielny don't work - because Koscielny wants to get tight, but doesn't trust SS to cover for him.
When Arsenal don't play well at the back, the majority of the time it's because Song, Djourou or Koscielny aren't playing (or Almunia is).
Together, Djourou and Koscielny have only conceded 0.66 goals/game, while the stats about Song speak for themselves, compared to Denilson:
Arsenal's win percentage with Denilson starting this season in the Premier League is 33%, without him 65%.
Stats courtesy of @Orbinho
Oh and also: Not only did Koscielny have the highest pass accuracy yesterday, he had a 100% tackling (5 out of 5) and clearance (4 out of 4) rate.