Yes, it is what I'm saying. However, I don't think even if the balance turned out to be positive on Akamai's side, even that would count as "asking ISP to pay for access".
Imagine a small ISP. Not a lot of hosted content. In order to boost local content, this ISP provides co-location services at lower than usual costs. Due to the same considerations, the ISP pays a lot of peering costs (mostly incoming traffic, not a lot of outgoing traffic).
And then this ISP has an idea: I'll contact Akamai. The Akamai network accounts for over 25% of web traffic. If I have a local Akamai presence, this will greatly reduce my peering costs. Akamai's sales people are aware of this equation, of course. As a result, the deal finally negotiated mean that the ISP is paying Akamai for the privilege of hosting Akamai servers!
And the ISP is ecstatic. Yes, they are hosting a commercial server for free AND paying for the privilege, but they are saving a bundle on their peering costs. This is a straight bandwidth for bandwidth money-equivalent transaction. If Akamai started asking for too much, the ISP could tell them to take a hike. Presumably, for that to happen Akamai would have to ask more than the bandwidth costs it is saving!
Should Akamai choose to play dirty (as far as I know, they never do), they would be in a stronger position than Netflix. After all, you can get Netflix content elsewhere. Conversely, you cannot get to, e.g., apple.com without going through an Akamai server. If Akamai isn't doing it, I don't think there is any danger of Netflix doing it.
Shachar