What's neat about Google's products, and the way they announce them, is that no one ever knows whether the product is:
a) A 20% pet-project developed by a couple of engineers in their spare time over a couple of beers, or -
b) Part of Sergey & Larry's master plan for world domination for which they're hiring 1000's of engineers and spending a shitload of $$'s on capex.
There's no confusion as to the significance of Vista or Live.com or adCenter for Microsoft, for example. But with Google's product you just never know and that is quite amazing when you come to think of it.
My take on Google's video ads? - If it is option A (20% pet-project of a couple of engineers), feel free to ignore all of the following.
However, if it is option B (part of the world domination master plan), then I think this is actually a pretty smart move. Not that I think the current implementation will work very well - I actually don't. As Mike Arrington pointed out, the incentive for the user to click on the video ad is questionable, there are no conversion opportunities, etc.
As I pointed out in my last post, the chicken&egg problem of any marketplace makes it very tough for insurgents to break through an existing leading marketplace and become a viable one themselves. So if Google's long term master plan is to become for TV networks what AdSense has become for online publishers, it will find it nearly impossible to do regardless of how great their technology for aggregating and serving video ads is.
The TV networks simply have all the ad inventory and the advertiser relationships and there would be no reason to use a great technology that's absent of real advertising $$'s behind it.
But that game completely changes if Google were to come to the TV networks (or TiVo & Co, which will surely get into the ad game sooner or later) with a war chest of tens of thousands of advertisers, and billions of $$'s in unspent budget. That solves the chicken&egg problem, and would enable Google to quite easily become the dominant ad platform for digital TV (not to mention taking over all advertising for the exploding online microchunked TV in the form of YouTube, iTunes and of course Google Video).
So using Google's existing asset (AdSense) to leverage them into the real target marketplace is a smart way to do it, regardless of how lame the current implementation on textual websites may seem. Microsoft has proved over the years how well criss-cross product leveraging works for creating monopolies in markets it didn't even play in, and Google seem to have taken notes. They did it in the past by leveraging their huge search user base to create a huge advertiser base (AdWords), which they later leveraged to become a huge ad network (AdSense), which they now seem to be leveraging to create a huge video advertiser base, which they will probably later leverage to become the dominant TV ad marketplace.
But then again, this may just be an engineer's pet-project... ;-)