I can see wireless as an issue -- where I live I share a tower with rush-hour traffic which leads to very slow LTE speeds during the commute times, streaming anything is just painfully slow at those times.

Still, charging the source isn't going to solve the last-mile congestion issues. At most it gives them extra "free" money to upgrade their network (or give bigger CxO bonuses) so they don't have to charge their customers more and risk loosing them to the competition.

If you are getting paid to deliver traffic to consumers then you would have a lot of incentive to not have congestion, since that would be losing potential revenue. That would if not solve at least improve the last mile issue.

I'm not sure it's that simple.

E.g. with WiFi just throwing in more APs may actually make things worse. Better hardware is a way, but up to a certain extent, then you hit the ceiling and need a new technology. Like the only working solution for a noisy environment is to just forget about the 2.4GHz band and go to the wider and less busy 5GHz. My (uneducated) guess is this is somewhat similar for stuff like LTE, too.

Not to be all "the market will sort things out", but in general if there is incentive to serve traffic there is at least a lot more potential to fix these things. If you lose potential revenue by congestion, you want to invest in more infrastructure to capture that revenue. Whether that is more base stations for LTE, wifi on public transport or something else. Today that is a cost that you have to recoup from the consumer.

My point was, for very-high demand situations, sometimes tech is just not yet there to support it. And new tech either requires lots of upfront investments (like, replacement of all the existing routers) or doesn't even exist (e.g. that 5G stuff).

I don't know how high the demand is, though. It could be hitting the ceiling, or it could be nowhere near that (and issues could have a different cause).