Just out of curiosity, what sort of calculations went into your result of "tiny additional costs"?
The reason that I ask is that bandwidth is finite and shared by all of the provider's customers. If one person consumes a lot of bandwidth, it will be to the detriment of the provider's other customers if the sum of all requested bandwidth is greater than the finite available bandwidth. Customers will eventually notice this, and the provider will lose those customers.
How did you account for this in your estimate?
You'd also be right if the typical overage fees were within an order of magnitude of what it actually costs the provider to ship those extra bits over the wire. Or even two orders of magnitude. But from the figures I read (costs to the provider of circa a penny per megabyte or well under that), we're getting jobbed.
Don't you find it a bit silly to constantly advertise "4G" networks and their supposedly blazing speeds, and in the same breath set up data plans which will only allow you to use your phone at full speed for circa 1 minute per month? Do the math yourself.
I agree that current data costs are astronomical due to an oligopoly situation (no real competition).
Using up 10x more electricity or water has very real, physical costs which are proportional to the amount consumed.
Using 10x more data has only very tiny additional costs to the bandwidth provider. A few seconds of research can easily confirm this. Asserting that the levels of bandwidth charges we're seeing right now are "fair" is hilariously wrong.