P.S. Try scaling up a 100 servers on the cloud using Chef, then try doing that with tools/platforms from 10 years ago... No don't. It'd be irritating as shitting on an ant hill.
That's ops. If that's what you want to do, more power to you, but someone who wants to code wants to code; and those of us who just want to code are EXTREMELY happy to have ops and dev-ops people around. Personally, I'm happy on either side of the fence; less so if on both.
And I'll admit to being in that group for bills, and things that look like bills, in the name (and letterhead) of a company I know I owe money to, beyond the first couple of invoices. The net result, for me, has been positive; much less information to remember and associate; just a name and total; so I can quickly and easily catch discrepancies and know which invoices actually require further investigation. The billing address from the first invoice goes in my address book (or I pay online, via a bookmark), so the worst case scenario is I overpay one of my accounts one month because someone sent a fake invoice. Just recently, I did receive a legitimate invoice from my ISP that included a billing error; I have autopay through them, so the payment was automatic, but the refund was also issued within minutes of calling them about it when I saw the abnormally large charge on my credit card. This event prompted an investigation on my end, wherein I verified the amounts billed by my utility and service providers over the preceding years, and matched up each payment sent (or appearing to be sent) to any of them with the payment information on file for each; my findings were positive, one billing error (easily corrected) in one year, across all of them. And I'm sure if I went back farther, that's all I'd find.
If you know who you do and don't owe money to, the "just write the check" method actually works quite well.
Game DLC could be handled the same way, with the DLC app unpacking content onto the filesystem when called on by the main app, which would remove the content when it was done with it (thereby freeing up space on your device by allowing the DLC to remain packed in, say, a gzipped tarball when not in use). And it wouldn't, technically, ne an in-app purchase, so it would skirt this new initiative.
You've got it right, but I'll go into a bit more detail for you, as well. In the connectivity market, there are basically two products you have to be familiar with: peering and transit.
When talking about peering, two networks agree to connect to each other and pass each others' traffic, assuming the flow of traffic is more or less balanced. If there is an imbalance, the network sending the additional traffic agrees to pay those costs, because they derive benefit from the other network passing their traffic along. In short, peering arrangements work because two providers would otherwise waste time billing each other roughly equal amounts, spending more in the process of generating invoices, processing invoices for payment, writing checks, sending checks, accepting and processing checks, and closing out the paid invoices, than any small imbalance that might occur. And there are other providers available through which these networks can pass their data in order to have it reach a given endpoint, but these routes are often longer; peering benefits both parties by providing both with shorter routes to some endpoints.
Transit is almost the exact opposite of this; and both Netflix and Verizon purchase transit from Level3. With transit, the customer is buying connectivity, much like you or I would buy connectivity from a last-mile ISP, agreeing to pay for all traffic across that transit link. The reason for this is that Level3 derives no benefit from passing traffic through, or accepting traffic from, Verizon's network; rather, Verizon benefits from the connectivity Level3 provides.
Peering only works when one network is passing traffic through to another network, be that other network an endpoint, or another peering provider. Verizon is an endpoint (as it Netflix, thus why both purchase transit), meaning that the traffic that enters their network terminates within their network and the traffic that exits their network originated there, as well. They are not a transit provider that can offer peering arrangements, they are an endpoint that must buy transit, but they are trying to behave as though they are a transit provider.
It seems as though you already understood the general mechanics of this; hopefully these details can help cement that understanding.
I don't think Netflix *wants* to fight dirty, or they would have done this already (and with Comcast, as well).