I think Siegler is missing the forest for the trees. The real question is not whether Microsoft should have dumped "legacy Windows"; in many respects, for ARM-based mobile devices, it actually has. The more interesting issues are around two industry-wide trends -- the increasing commodification of software, and the re-verticalization of software and hardware -- and whether the Windows business model itself might be in danger of obsolescence.
Also worth noting: for Windows Phone 7, Microsoft did actually ditch backwards compatibility in favor of "the new thing". And despite generally favorable reviews, it's not exactly burning up the sales charts.
It was easy for MS to ditch compatibility for Windows Mobile. It never was as dominant as the PC share so they could afford to start with a clean slate.
In terms of ARM tablets, I assume that the only reason to have desktop mode is so it can run Office, however limited the features may be since it would take much too long to make it Metro/touch-friendly. MS isn't completely cutting the cord like Hastings is doing.
I love reading MG Siegler, Gruber and company but all this talk about how Microsoft should emulate Apple's strategy leaves me sad. We should be pushing for innovation, not copycats. Microsoft is trying to pull something hard but they might actually make it work, why do they need to just copy Apple's strategy just because it has been working for Apple? I have said this many times before but there's not only one way to achieve success.
Netflix is moving too fast in response to their subscription splitting. "Quikster" (sp?) looks like an overreaction to a brief, albeit significant, customer irritation. Rebranding the core product (it's long-term fame is for DVD rental, not streaming, even if the latter is the main goal) confuses people. "Hey, where did Netflix disc rentals go? Guess I gotta go to Blockbuster again..." I don't see this ending well.
The kind of moderately tech-savvy person that's Netflix's main customer is going to completely abandon physical media in the next year or two. Since getting Apple TV (i.e. iTunes movies plus Netflix) I'm no more willing to wait two days for a disc to arrive than I am willing to wait for a search result. Movie watching is often social, and more and more people are accustomed to having an immediate say in the movie they watch in social situations (probably thanks to Netflix successfully integrating into every TV/game console/set-top box).
So the downside to abandoning physical media is limited (they'll still own what's left of the rental delivery business) and the upside of staying in the lead of an increasingly competitive industry by focusing on on-demand video is huge.
The "long term game" is whatever Netflix defines it to be. Success at this game is only ever a combination of clear vision matched with execution. Hastings has defined their end-game and made that vision explicitly clear. He has freed the management teams and their whole organization to now pursue that wholeheartedly.
Premature scaling and optimization are the top killers of startups. A better title might be "well established companies rarely die from moving too fast". Those companies have a larger safety net and can afford to take more risks, because those risks are small to them. The same risks can (and likely will) kill a startup.
Companies rarely recover from moving in too many directions at once. People forget that a company isn't something magical. It is just a name on a piece of paper.
They may not have died, but I think Microsoft suffered from moving too fast in the mid-late 90s: the antitrust trial, security and reliability problems.
Also worth noting: for Windows Phone 7, Microsoft did actually ditch backwards compatibility in favor of "the new thing". And despite generally favorable reviews, it's not exactly burning up the sales charts.