Widgets vs. native apps
Only it's not.
Not exactly. Raby's right to say the web is a key element of making mobile apps work cross-platform. He's essentially talking about cloud-based services. And while widgets and web apps generally have a patchy track record in terms of performance compared to native apps (depending on what the app does - for example, RSS reader widgets run well in the cloud, navigation maps less so), arming them with network APIs could help, along with things like faster connection speeds and 1-GHz processors.
But as a "write/submit once, deploy/sell everywhere" fragmentation killer, it doesn't accomplish that much. Only one OS platform player - RIM - had pledged support for WAC at press time, and regardless of how many follow suit, none are going to give up their app store strategies, not even as mobile content migrates to the cloud. Apple most certainly won't - its legendary restrictions over apps development are partly the result of Steve Jobs' belief that a standards-based but tightly controlled OS ecosystem is crucial to surviving in a cloud-based content future.
And however long that future takes to arrive, the market for native apps and the app stores that sell them will thrive in the meantime. WAC's Raby admits as much, saying during the press briefing that widgets and apps storefronts will co-exist in the same way that web games co-exist with console games.
None of this to say WAC is a bad idea. If nothing else, it will make billions of feature phones eligible for the apps craze, which developers will likely appreciate in terms of access, scale and approval processes (as long as WAC is sensible about it, and provided Nokia, Motorola and HTC support it). But the mobile OS landscape will remain as fragmented as ever.