Morton's Fork of DRM

Derek Slater is pondering DRM again:

I feel like there's a huge divide between people who think that DRM has no chance because consumers will always be able to get digital media for free, and the people who think DRM has no chance but consumers will still pay.

Within that first category, there's another subdivision: people who think that means we need DRM that's far more difficult to crack (TCPA, Palladium, etc.), and people who think we need a more radical solution (Professor Fisher's plan, for instance).

I don't know which side is right. But, I think a lot has to do with how good DRM can be and, perhaps more importantly, how good do we need it to be. If all DRM is crackable or can be gotten around via the analog hole, how many people do we need to still purchase things legally to remunerate creators?

(You can also take the position of Kevin Marks, who says that DRM no matter how good is doomed to fail, because it "destroys value." I'm not so sure of that, inasmuch as, if reasonable restrictions could be built into DRM, then perhaps consumers would accept it. I also don't know if such reasonable DRM could exist.)

That's my Morton's Fork of DRM.
It is always going to be crackable by emulation, via the Church-Turing thesis, or by analog means, so a determined person will break it.
The only reason for applying it is to make files less convenient to customers, so it automatically makes them less valuable to them.
The BALANCE act should go a long way to making DRM 'reasonable', but as that point approaches, it is even more reasonable to avoid the entanglement altogether.