I just tried Zumobi. It sucks. Now, I don't like any mobile widget platform, actually, so that isn't sort of a fair summary if you want to compare this with other mobile widget things out there like Nokia's Widsets, but even so Zumobi isn't a shining example.
* The widgets you can add so far are either news feeds, or ponderously slow applet sorts of things (tip calculators anyone?). Where do the widgets come from? Their online gallery of course! And if you're a developer, how do you get your app in the gallery? Well, you need to use their toolkit, and then submit your apps to them for "certification", and I guess they might then add them to their site ("Contact Zumobi Developer Support when you're ready to submit your tiles for certification.") Yes, it's as bad as it sounds.
* The zoomable interface is a joke - it's pure eye-candy and overly complex considering the fact that you have only 16 tiles you can fill in the first place which fit pretty easily on one screen without zooming, and when you add/remove widgets you can see that. The zooming doesn't even work well or look nice - it's slow, and gets way too pixelated.
* These guys seemed to have hired PR and Marketing folk before any competent engineers judging by the number of articles I've seen about it lately, because the coverage definitely isn't because anyone's actually tried to use the app. (Note to the folks at Zumobi... your definition of "open" and everyone else's are extremely different - which considering your Microsoft heritage is not surprising.)
Beyond the technical details, I personally don't understand the reasoning or business model behind an effort like this. They're not charging for the app (which is good, as no one would pay for it) and assuming they're not going to charge developers for certification (which is good, as no developer would be that stupid) they must be expecting to make money from the banner ads that take up 20% of the non-scrollable screen (the demo apps had Amazon and Astrology.com mobile ads). In order for that to work, they would need insane penetration of users, and you know what? Mobile installable apps don't get that many users, especially targeting just WinMo phones. This sort of thing might work for one of the big portals - MSN/AOL/GOOG/YHOO - because it could be some sort of affinity play, get users to come back to your main site, bring along your email, etc. But as a standalone company? Even if the widgets were useful - which they're not - there's no way they can be profitable with this thing. The numbers just don't add up.
This is definitely one of those companies where you go in and ask the employees, "Hey, how many of you used the product on your phone today?" and you'd get this sort of response: Sales guys would have it on their phone and know only how to start it, the dev guys would be doing some testing on a version that won't see the light of day for a long time, the execs would have a version on it from months ago which they used to close funding, and anyone else wouldn't even have a phone it could run on. I'd bet money on it.
One last thought... I wonder if these guys will get a .mobi domain? Just for fun... like zu.mobi... That'd rock! It wouldn't help the product suck any less, but it'd be a cool domain. :-)