I think their big fear is people buying a hackable version and setting it up to be a generic epub reader instead of a walled-garden sales point, thus losing money on the hardware or perhaps support costs for people who messed up their devices. But hacking hardware, even if it just means transferring a .deb and restarting, is utterly beyond the kind of people who would rather just press "buy & download" and forget about it, which of course is 90% of the market anyway.
In either case a premium hackable ereader would be an interesting (though perhaps not profitable?) direction for anyone, even an indie company, to pursue.
It is simple economics - as much as us (engineers, developers, etc) believe we are a large market, we are not. The overhead of creating a different SKU to serve <1% of a market is not enough. Besides, the money is not in hardware, it is in content.
The first thing I do when I buy an Amazon ebook is get rid of the DRM. If I couldn't, I wouldn't buy it.
Don't underestimate the appeal of hackable.