Perhaps then the goal is to have a mutable pricing structure that responds to market pressure. If it really doesn't cost anything to produce 1 additional electronic copy, something like this might work.
At launch date / first to market, there is a high price tag for the digital good (whether it be music, game, movie, etc). This is so that the vendor/author can make sure that they can sell at a high price to those "first movers" that actually want to pay $1000 to be the first person on the planet to have that special ring tone, ahead of everyone else in the market, for example.
As time passes, and as more people buy the "product" (which is done via a download or a file transfer) - the price of the product goes down - because the as demand rises, the cost of production actually doesn't change - and the author already made some money, so there's no point being greedy, because then people would go and pirate it.
Once this price goes down, it goes down for EVERY SINGLE customer, including those who bought it during the high initial price. I mean, the vendor does have the customer's credit card numbers. Do a residual, scaled, periodic refund for every single customer, based on the differences between the current discounted price and the original price at which the customer bought the good.
As the # of units sold increases, the price of the product decreases - of course the relative rates can be tweaked so that the vendor is still making more money than they are discounting, because as an electronic/digital good, there are no factories or warehouses - it's easier to make money at mass scales when it comes to digital goods (vs physical goods like bicycles).
Existing customers feel validated - the more people they recommend to buy the product, the more money they get back.
New customers don't feel like they're getting ripped off - the prices are falling all the time, and you "buy into" a whole discounting structure.
The original vendor/producer can finely tune the ratio at which the prices fall in order to control profitability, popularity, and sales rate as a factor in promotion and marketing.
Again, if you take advantage of the fact that if you sell 10 copies, the cost for building the software/digital good is no different from when you sell 100,000 copies, then you can exploit that "efficiency gap" if you can call it that between the existing pricing structures and a variable pricing structure that I describe.
At some point, you will reach a "price floor" at which the producer is comfortable reaching their target cumulative total profit. Since altering the price changes the price for every single customer since the beginning of the lifetime of sales for the product, you can actually say - I want x million for this product in 3 years time - and actually be able to hit that target by planning out the way the price changes. OR if you believe in the long tail, go ahead and have a constantly decreasing price - as you reach the hundreds of millions of people in the long tail, you can charge 1 cent - and you would get hundreds of millions for a product you made - providing usefulness for all those people at a great price - AND being able to protect your livelihood during those critical first few sales.
Of course, this requires every single sale to every single customer to have its own merit, its own tracking system, its own values. BUT personalized sales is exactly what we are looking forward to in the future.
And I think technically, it's not super hard - I mean we have nonstop systems and mainframes and clusters of all kinds - that analyses the trillions of collisions in the LHC - what makes us think that we can't track six or seven billion consumers who are alive on this planet, who might be making 5 or 6 million transactions a year? It's not that bad when you consider the scale that google / national science foundation / CERN operates at.