Further, there are tactical reasons why a change in high defenses matter more. Many buff/healing powers can ameliorate an effect; but these are usable only a limited number of times. For instance, a halfling's second chance is worth more when the reroll has a high probability of missing. Many leaders can remove a save-ends effect, but can do so only once (or some limited number of times) - so these powers become a more effective stopgap when the base defense is already good.
This is the singular point I can agree with - Second Chance is more effective when your defense is high.
Given that increasingly high defenses can result in you not being hit by a status effect at all, your argument about ameloriating effects only being available a limited time does not justify increasing higher defenses more. If anything, it suggests increasing lower unless all of your defenses are sufficiently low that you're going to get hit by that status effect multiple times (which is totally possible).
Effects do muddy the waters: but I'm convinced that for (most) effects too, changes to low defenses matter less than changes to high defenses.
I believe that you're correct in two very specific cases:
1) Your defenses would be hit on a number lower than a 2 if that were possible (extremely high level, or DM/module using monsters too high level)
2) You have a plethora of make an attacker reroll an attack powers
Which I suspect are not actually the reasons you're selling.
Let's take a theoretical 10 round combat against a solo in which every round it does a stun attack on two people.
The first is hit on a 3 (90% chance to be hit). The second is hit on a 11 (50% chance to be hit).
So, the first is being stunned 9 of 10 rounds at the moment. Which sucks, oh boy. The second is being stunned 5 of 10 rounds at the moment. Which still sucks, but is significantly less annoying.
If you gave both +4 to the defense in question, the first would no longer be stunned in 2 rounds. And the second would no longer be stunned in 2 rounds. Absolute effect, identical.
This would change their rounds of usable combat from 1 to 3 (+200% rounds) or 5 to 7 (+40%) respectively. Comparative effect, to the lower defense.
In terms of effectiveness, the first guy is now unstunned long enough to get to throw out his strongest encounter powers. The second guy gets to throw out two additional at-will powers. Comparative effect, to the lower defense.
I'm not entirely sure how you'd quantitate fun between the two - but I can say that I suspect being able to act two more times would probably remove a lot more of the frustration from the 1/10 guy than the 5/10 guy. Purely cause he'll have a lot more frustration built up. So comparative effect, to the lower defense.
So, simple example in which improving the lower defense was more useful. And there are lots of others. The warlock or ranger who gets dazed less so they can quarry more, which is hardly something they need to do every round. The fighter who gets immobilized less so he can get into melee then lock something down. These are things where being near autohit is potentially crippling and being able to avoid that for just a few rounds is a fantastic boon.
I don't actually advocate boosting lower over higher - I'm more disputing your claim that higher clearly over lower.