Scalzi, who already gets more publicity books than he can read -- hell, I get more publicity books than I can read, and I bet Scalzi gets more some days than I get in a week -- was not having any of that; he made it very clear he wasn't going to jump through any hoops to read a book, and complained loudly about the practice on the Internet.  As usual, he got a lot of comments, the majority of which were solidly on his side -- though why anyone who doesn't review books would care about the mechanisms of the process is beyond me.
I've got a somewhat different perspective on the issue, though: one forged by being one of the few people on both sides of the equation, as both a marketer regularly seeking publicity for his books and an online book reviewer looking for interesting stuff to review.
Edit, later that day: see comments; Scalzi did not say or imply that reviewing a book is a favor, and I mistakenly attributed that mindset to him. So it may be that I'm agreeing with him at the top of my voice, and I apologize for that. But many Internet reviewers do seem to think that a review is a favor -- though they're divided on who is giving a favor to whom -- so I'll leave the below argument intact.
Certainly, any books that come to him asking for blurbs are requests for a favor. But the books he's writing about this time are much further along in the publication process; they're well past blurb-gathering, and are aiming for publicity on publication. Scalzi's Whatever blog is at least semi-professional journalism (depending on how much weight one puts on the question of getting paid), and pitching coverage to a journalist is one of the central activities of a publicist, and doesn't generally involve a favor on either side. Sure, Whatever isn't Time or Oprah, but it's a regularly updated public media entity that gives regular coverage to specific projects and has a systematic policy for dealing with submissions. It walks and quacks like journalism from where I sit, so I would say that it's entirely a matter of business -- no favors asked for or received.
On the other hand, journalists are busy people -- both the traditional kind who get paid, and we odd Internet loudmouths that generally don't -- so throwing several obstacles in their way is not a good idea. I review comics, and I've seen a number of products as PDFs -- a great idea, since printing color ARCs would be ruinously expensive, and I'd be so far down the publicity list that they wouldn't make it to me anyway. Those PDFs generally are gatekeepered -- with passwords or secret FTP sites or something similar -- but are not otherwise encrusted with DRM, and that strikes me as a good compromise. The works aren't out there, freely available to anyone, but -- once they get into the hands of a reviewer -- they're easy to access and use.
That policy isn't particularly scalable, though -- it does require a gatekeeper on the publisher's side to give each reviewer access to the secret FTP site, or a password, or whatever. And so publishers that want to get hundreds of electronic copies out of each of hundreds of books to a vast array of media folks (including we shabby blogging types) are likely to want a mechanism that doesn't require as much tinkering. And Eos is definitely using that model; they're looking for mass Internet publicity (what we in the business like to call "buzz") for a number of books.
In my experience, publicity opportunities are tiered. The tiers are slightly different for different projects, and some top tiers (Oprah, The Today Show, New York Times Book Review, etc.) are not even plausible for most projects. But there's always a hierarchy, which is similar to those of a college-bound teen: the stretch outlets, which might come through and would mean a major publicity boost; the middle tier, which are pretty dependable and reasonably strong; and the safe hits, that cover nearly everything they see but don't move the needle much in views or sales.
Most blogs are in the third category; I certainly am. Whatever is in the middle tier for most projects, possibly ranging up to the first level for certain kinds of SFF books. And that's where the balancing act comes in -- top-tier outlets are generally more important than any project they're considering covering, while bottom-tier outlets are notably less important than anything they'd see. And so publicists develop different methods for dealing with different levels -- they're much more deferential towards an Oprah producer than to a blogger, and spend vastly more time honing a pitch and making repeated contacts with the former than the latter. There's a power relationship between publicists and the media, and both sides can get defensive about it -- particularly when they are assumed by others to have much less power than they think they have.
So big "gets" command a lot of individual focus from a publicist, while smaller media are blasted with a mass mailing (whether e- or snail) -- targeting increases the higher up the food chain an outlet is. And the deference a publicist shows to a media outlet, and the level of effort the publicist is willing to expend, increases greatly as that media outlet moves up the tiers.
That works well...as long as both sides agree on the relative importance of a media outlet.
But it can lead to problems when traditional publicity meets web culture; many publicists are very web-savvy and follow Technorati/Alexa rankings, but some others, frankly, don't really know that certain web outlets (again, this varies by topic or genre -- Will Wheaton is an Internet god, but he probably won't be much help with a quilting pattern book) can be huge for them...and that the people running those web outlets know that, and expect to be treated with the respect due their influence.
So, buried in Scalzi's rant, is a sense that he's a tier-one (or at least tier-two) outlet being treated like the rabble in tier three. And a few of his commentors are tier-three people who have noted that they've jumped through those hoops themselves, and didn't really mind it. Eos is treating him like general review rabble; they should have pulled him out of that list and given him more attention and thought.
On the general question -- are heavily DRM-ed e-ARCs a good idea? -- I'd have to say that my answer is a solid maybe. For mass-marketed fiction, particularly books in currently popular genres with strong Internet communities and a wealth of potential semi-pro review blogs, it can be a great way to get a book out to a large audience of tastemakers. But it shouldn't be the only strategy; those top-tier review outlets (on print or in pixels) will be offended if asked to jump through those hoops. So there needs to be a different strategy for the more important media.
It's difficult to say from here whether Eos has a tiered strategy like this -- I certainly hope that they do, that they're not trying to send a paper link to a DRM-ed ebook to Publishers Weekly, but publishers have done plenty of dumb things before -- and so it's unclear if their mistake was in having a single low-level strategy or in wrongly assigning Scalzi to the lower tier of a larger strategy. But, in general, publicity is the art of making your target media interested in your pitches and happy to see them...so this most certainly was a "publicity fail."
Appendix: One small side issue -- the reason publishers DRM files like this is because semi-insiders and early adopters are much more likely than average consumers to be the source of pirated copies. (My company recently studied the issue, and found that nearly every single file of ours on torrent sites either came from page-layout files or pre-publication review copies.) So any cries that publishers just need to "trust"reviewers will be met with a haughty laugh and a shake of the head; that's been tried already.
 Via UPS, of course -- I've noticed many times for myself that certain publishers seem determined to send physical packages in the most expensive way possible.
 Digital Rights Management -- software designed to limit the uses that someone can make of a digital product. Content providers like it, because they want to be able to define what they're selling or giving away, but audiences that actually know what it is tend to uniformly loathe it.
 And what is the Internet for if not for complaining, anyway? OK, for complaining, cute pictures of cats, and porn. But other than that?