Erm, are you saying it is wrong for it to be based on that?
Say I illegally download a song and pass it on to a friend, who passes it on to another friend, and so on, until there are 10 people who have the song.
RIAA's argument is that the damages are the 10 songs existing on 10 PCs, and the 9 times it was shared. They also say that the 10 songs are a lost sale.
However, say the song is Rebecca Black's new debut album. It is utter shite. RIAA's argument is therefore "if the song wasn't shared, then we would have got 10 Rebecca Black CDs' worth of revenue".
This is clearly wrong - you aren't going to get 10 Rebecca Black CD sales out of the 10 people here - the CD is crap! Just because the song was illegally obtained doesn't mean the song
would have been legally obtained if the illegal version is not available.
Even though piracy is not theft, thieves stealing a Ferrari does not mean they would buy a Ferrari if they could not steal one.
The actual damage is the people who
would have bought the CD if the illegal copy wasn't available - and it can only be approximated, of course. But it implies that the damages should be equal to "expected revenue minus piracy" minus "expected revenue with piracy" - in other words, it has to be less than the expected revenue minus piracy.
(And it assumes piracy always has a negative effect - for an independent, struggling singer or band, it could well be better if the song was mass-pirated since it spreads the song's popularity, indirectly boosting revenue.)
(Also, there's the murky fact that those who pirate more music are more likely to buy music...)
(I accept this doesn't always work for all industries like the film and video games industries - in the latter case, it can really hurt - but for the music industry, is it really hurting things?)