So, it's to be a battle of wits, is it?
First, I gather you know much of the US has nowhere to go if they want to switch, making your first point moot. But, I fail to see how it's any different from any other business doing cost analysis based on estimates.
Reasonably populated areas in Europe pay far less than reasonably populated areas in the US. New York City and San Francisco, two big cities, have some of the worst service in the country. Why? Effective regional monopolies. NYC is suing Verizon for failing to meet terms. Verizon doesn't want to build more. Can't imagine why. (Simply, costs more to maintain than wireless. Why serve customers when you don't have to?)
And yes, I carefully worded that because I'm not a network engineer for a major ISP and don't know how they over-subscribe. I thought I made it clear that is acceptable practice but if not: that is obviously acceptable practice. Theoretical peak is just that: theoretical. I'm not going to make a case for a restaurant to build out for unused tables and will not make the case for an ISP to build out for unused bandwidth. Both are silly. Again, I thought that was clear. Haha, wasted your time.
The argument is ISPs are, in fact, super-villains. Why? Data caps have nothing to do with over-subscribing or bandwidth management. I'm not comparing this to restaurants because the analogy would be poor. In fact, none of what you argued have anything to do with the reason most people hate their ISP. Bits are not consumed. The only scarcity is bandwidth. Restaurants have a max capacity but that does not cap the speed at which people can be moved in and out. People would be pissed if they were rushed because they had a time cap and would go elsewhere. Whoops, I did what I said I wouldn't do. And yep, the analogy is awful.