This is going to have to be a multipart comment, I wrote so much:
1. A cure for the disease of which the RIAA is a symptom.
My idea for this is included in a more robust solution for #17.
17. New payment methods.
Here's the $100 billion dollar idea (if it could be pulled together). Make a new credit card (or work with an existing company). Consumers would sign agreements to exclusively use that card (more financial info could be asked for, though). In exchange, the company takes all the data about the purchases, and any other (anonymized) info the company can gather about the customer's financial situation and habits (the more info the customer gives the company, the better the prices they are given), and uses artificial intelligence to make predictions about how much a person would be willing to pay for something. In exchange, the customer agrees to accept the price the credit card company charges, without knowing the price ahead of time (not to be higher than the listed retail price). The retail establishment chooses the minimum price they will accept.
Example: a customer buys a room at a fancy hotel, which would normally retail at $400 per night. You know that the customer never pays more than $150 for a hotel room, and there don't seem to be any other indicators that this is a special event where the customer would pay the full retail price. Your AI estimates that, based on his income, recent expenses, and past hotel habits, he would be unwilling to pay more than $200 for the room. The minimum the hotel will accept is $150. You charge him $190.
The customer just got a great room for $190, and as long as on average you save him money on his purchases, he's happy. The hotel just made an easy $40+ extra profit for a room that would have gone unused (if they were closer to full capacity, they would raise their minimum acceptable price). You charge a % commission, and also have the exclusive card of the customer, worth a lot.
As long as the predictions are reasonably accurate enough in the long average, everybody wins. Hotels, airplanes, buses, sellers of copyrighted goods (music, movies, books, software, games, etc.), and any goods sold with a sizable profit margin (prescription drugs, etc.), make more money by selling more of their "excess capacity". As long as the average price per unit sold is not too much lower, relative to the extra number of units sold, they win. The customer wins with goods and services that are never more expensive than retail. And you win by charging a %. This is basically a technological solution to the huge problem of "how to capitalize on every strata of the market, for the same exact thing".
If someone tried to buy stuff through a "cheap buyer", the data would quickly mount up on that, and you can call fraud (or retroactively charge them more for everything they've bought recently, since they are now demonstrating evidence of heightened income- that'll be in the contract).
Based on my past experiences with AI, and what I know about AI as it's already applied by credit card companies (for fraud, etc.), and the kinds of AI done with "store loyalty" programs, all of this is quite doable. It just has to be pulled together.
The only problems I've been able to find with this idea are that a) I'm an environmentalist, and don't care to see people able to buy even more junk (but maybe they'll spend more time on stuff like music, which isn't bad for the environment), and b) some rich people might get annoyed that others are systematically paying less for the exact same goods and services as they are paying for. But people are used to poor people paying less for education, health care, etc., so this might not be a big deal.
c) if you think that sellers might significantly raise the retail price, to try to make more money off of the upper part of the customer strata, you're wrong. Anyone that tried to do that would be undercut by competitors who would be willing to sell at a normal retail price, to win over all the high-dollar buyers. That would still be the biggest part of the market (since it's around 90% of the current market, and the market for people willing to spend less than retail is obviously a limited market). The only sellers to watch out for are those with monopolies on crucial products, since it would be hard for other sellers to indirectly compete with them.
But is there a solution that is simpler and targets well to music, movies, etc.? What about selling, like, a contract for 100,000 songs? The buyer provides a little information, and if they look like they couldn't afford that many songs individually (college students, etc.), they get a flat rate to download 100,000 songs of their choosing (maybe $10,000, paid in installments over many years, if they're a college student). Obviously few people could afford or be willing to pay 99 cents per song for 100,000 songs, so if you can do a decent job estimating the maximum music budget for someone, just charge them that rate and give them everything they could possibly desire (which would also solve the problem of paying 99 cents for something you never listen to after the first spin).
The main problem with this is security- making sure those 100,000 songs don't get ripped off. If the music companies could make security that actually WORKED, I think it would be a pretty good idea. Even most people who currently download stuff for free would pay at least a little for hassle-free downloads of copies they know will work.
There is a governmental solution, also, which I will go into later. It would be much better for the overall good of society, I think, and it would be the simplest way to go. It's just not the commercial solution I'm guessing PG is looking for.