Consider the people sitting around the table when a company like Microsoft or Apple decides to implement a DRM scheme. There are the developers and architects that just might understand how difficult the problem is and how brittle the eventual implementation will be. Then add accountanting, marketing, and the voracious negotiators from the content providers. Stuff everyone’s agenda and the technology into a Blendtec (http://www.willitblend.com/) and - voila! - a gallon of warm, soupy $#1t.
Once they have your money, nothing matters to them. You get stuck with the contents of the Blendtec.
If XBL is such a value-add, why not “fail open” with the DRM? That is, if the auth server is inaccessible, then assume auth succeeded and let the user play. Sure, someone could (conceivably) pirate a game and unplug the box from the network each time they wanted to play it, sort of like rolling back the clock to use time-limited demos. It’s inconvenient and retarded, but hey, if you have the time and attention span to do that, have fun - chances are you’re not going to buy anything anyway.
However, anyone who maintains online services knows there are a million reasons why a random user can’t get to a service - their hardware or software broke or got misconfigured, some jackass started digging without calling the 1-800 number first, any of the ISPs or NSPs between them and XBL dropped the ball, or the nice people running XBL got a bad batch of crack and turned off their pagers to watch “Beauty and the Geek”. The service will go dark for some number of customers, hardware will fail, especially first-generation machines. Users will be affected; how they will be affected is a software design decision.
It’s not like these problems are new. Think of electronic door locks as a form of digital rights management - a computer decides if you have the right to go through a door. How does the system respond if there’s a fire or loss of power? Do you lock all the doors and leave the people trapped inside to fend for themselves a la Cocoanut Grove (http://en.wikipedia.org/wiki/Cocoanut_Grove_fire) or do you accept the risk that some people will leave without paying their bill while the system is down?
Designing a robust DRM system that consumers perceive as fair is a herculean task; doing so under the direction of nontechnical stakeholders is nigh-impossible. What you end up with is a trainwreck that rewards the greedy at the expense of the unfortunate, a thoroughly 19th-century situation.
I can almost understand Jeff’s attachment to Rock Band; Katamari Damacy kept me sane through the waning months of 2004 and reintroduced me to console gaming after a 15 year hiatus. Still, he gets everything he deserves by buying into the DRM paradigm - heads, they win; tails, you lose.