In the net neutrality debate, a rarely mentioned aspect is the widespread practice amongst ISPs of over-subscribing bandwidth. Before streaming video became readily available and popular, this was viable, even reasonable, since few users managed to actually saturate their links for extended periods of time. Now that customers are suddenly demanding to get what they were sold, the ISPs find that they simply cannot provide this while keeping current prices.
To maintain the illusion of low prices towards the consumers, they instead prefer to make the major content providers (e.g. Netflix) pay part of the cost. For Netflix users it doesn’t really matter much how the monthly bills are split between their ISP and Netflix. It is, however, unfair towards smaller content providers unable to enter such agreements with ISPs.
The shifting of costs onto content providers is sometimes defended by saying that users should not need to pay for bandwidth they do not use, and the ISPs raising their prices would force users to do just that. While this is true, the argument conveniently ignores the obvious solution. The ISPs simply need to (re-) introduce tiered subscriptions wherein the cheap options have bandwidth caps (either rate-limited or with monthly usage quotas) and higher-priced alternatives have fewer (or no) restrictions. Metered billing is also a possibility.
Any move along these lines by an ISP would of course cause furore among high-bandwidth users who would see their costs increasing, so nobody wants to be the first to take such a step. Even ISPs in monopolistic situations would probably rather avoid the bad press, especially if they can extract the money from other sources (Netflix).
Strong net neutrality legislation would restrict the leverage available to ISPs seeking to collect payment from content providers, forcing them instead towards a business model of selling their users access to the (entire) Internet rather than one of selling content providers access to their users. This is a necessity for the Internet to retain the open nature which allowed it to grow in the first place.
Somehow this seems to be mostly US problem, in other countries ISPs somehow manage.
And I had enough experience with an ISP you describe: a local Ukrainian ISP that connected lots of houses with 100 Mbit/s Ethernet but outgoing channel was only couple of Mbit/s IIRC. First they had packages (i.e. 256 megabytes of Internet traffic), then they made those packages expire at the end of month, then it tied access to local resources to daily Internet usage—i.e. if you use less than, say, 10 MB/day you should pay for every megabyte you download from servers located in the local network. And that’s the only advantage of such local ISPs—local DC hubs (with content of dubious legality of course), gaming servers and such. I simply cut the cable at that point and switched to some saner provider. I don’t like getting headaches trying to calculate how much I need to pay.
Everything Comcast, AT&T, Verizon, Time Warner and the other large US ISPs are doing is all intended to do one thing and one thing only. Its intended to make ditching pay TV in favor of internet-based content options as difficult (or expensive) as possible.