There is something to be said, true, but I think TOO MUCH tends to be said about it with respect to Perk. Again, we now have years of evidence about how our defense responds with and without Perk. Essentially, his presence or absence just doesn't play much part overall in our defense. Yes, the things he provides are nice...but they're also replaceable.
I think we have evidence that he's less important to the defense than KG. The year before KG came the defense was just over 3 points better when Perk played. Last year, when KG wasn't always at his best, the defense was again just over 3 points better with Perk playing. I don't have stats for this year but I would suspect that our defense has been close to that much better with Perk in. If you keep track of our defensive efficiency at all our numbers seem to have dropped by a decent amount recently and I don't see any other major changes that would cause it. Perk's not irreplaceable, but it's going to be a tradeoff. We'll be a little worse defensively without him and possibly a worse rebounding team but we'll probably be somewhat better offensively, and our bench would get a boost from Green.
I see your point, but I think it needs to be developed a bit further (in other words, I'm about to use a lot of words and numbers to prove you correct, then reiterate my own point

). It is interesting to look at Perkins vs KG when looking for defensive impact, because I think it actually brings home both of our points. Yes, Perkins does have some defensive impact (you'd expect him to, as he is a good role player). But it's a very small impact when compared to someone like KG, and ultimately his scale of impact is replaceable. Let's take a look at some things (all of the numbers I'm about to use come from either 82games.com or basketball-reference.com):
From 2007/08 - 2009/10, the starting unit of the Celtics (Rondo/Ray/Pierce/KG/Perk) gave up
97 points/100 possessions in 3305 minutes (while scoring 112 points/100 possessions).
Over those 3 years, the starting 5 with
KG playing but someone replacing Perk (Posey, Baby, Scal, Powe, or Sheed) gave up
101 points/100 possessions in 656 minutes.
Meanwhile, over those 3 years the starting 5 with
Perk playing but someone replacing KG (Posey, Baby, Scal, Powe, Sheed) gave up
112 points/100 possessions in 914 minutes.
This (2010-11) season, through 57 games the Celtics have a team defensive-rating of 99.8 (1st in the NBA). Through 1/19/11 (the last 82games.com update), the Big 4 with either Shaq or Baby have
allowed 101 points/100 possessions (while scoring 118 pts/100 poss) in 418 minutes.
Conclusions: These numbers generally support the point that you were making, BballTim. You point out that the year before KG arrived and last year with him weakened, the Celtics were about 3 points better with Perk than without him. The 82games.com 5-man data shows that from 2007/08 - 2009/10, the starting five gave up about 4 more points/100 possessions without Perk than they did with him (101 pts/100 poss, instead of 97 pts/100 poss), and those exact numbers are holding true thus far this season. Would seem to indicate that Perk's presence was good for about 3 - 4 ppg on defense to the starting line-up. Our stories agree.
This is a reasonable defensive impact for a solid defensive role player, which Perk is. But the thing is, it's not a BIG effect. As you see from the same analysis, the defensive drop-off without KG was about
15 points/100 possessions for the starting 5. Again, the point here isn't to make this about KG vs Perk, it's to show the difference in SCALE in their impact. It's not just that Perk's defensive impact isn't as big as KG's...it's that it's NOWHERE NEAR as big as KG's. The 3 - 4 points that Perk might be worth on defense, those are points that can easily be made up.
We've seen this year that in the 418 minutes that the starting 4 have played with either Shaq or Baby the offense has scored 118 pts/100 poss and allowed 101 points/100 possessions (compared to the 112 pts scored and 97 pts allowed per 100 possessions that the Big 4 + Perk line-up produced over the last 3 years). Those 4 points lost on defense have been made up with 6 points gained on offense. Which, at the end of the day, is my main point.
Perk is a solid defensive player. You can see a definite, measurable defensive impact from his presence. On the flip side, he's not a defensive star. His impact, relatively speaking, is small on defense. And can easily be replaced. Yes, there's a bit of an offense/defense trade-off, but it's on a small scale. You're still talking #1 defense in the league with or without Perk, but instead you're looking at a better offense. On the whole, again, Perk's impact is replaceable.