From what I saw, hackday was used to pretend the organisation cared about the ideas of it's engineers and web developers, but at a safe enough distance not to affect their product roadmap. When push came to shove we were all enthusiastically encouraged to patent our ideas, because it protects us from evil people stealing all our ideas. Naturally I refused.
The Yahoo Open Hackdays also, was an exercise in futility. Yahoo didn't do anything with ideas generated, or encouraged the people who attended and built something. It was just a two day event over a weekend where a venue was populated by groups of people who may be building something, or not. Esssentially it was just a marketing ploy to get developers using the Yahoo APIs for a weekend.
Looking at that, these hackdays offered no value. I don't need a hackday / hack-weekend to work on ideas I consider interesting. I don't need to be in a noisy drafty building where the roof opens up during a thunderstorm ( http://developer.yahoo.com/blogs/ydn/posts/2007/06/hack_day_... ).
I gave up on the internal hackdays, after participating in two: winning one, and getting a notable mention for the second. Ideas are no good gathering dust on the shelf. And all hackdays accomplished was for the company to have an inkling on the ideas I was working on in my own time, so they could decide to claim ownership if they liked.
Open Hackdays are only useful to meet up with people you haven't seen for a long time. So having free beer and pizza is useful, I guess.
To be honest, the things I did gain from being at an open hackday:
* Watching James Aylett build a social network from scratch, and fully unit-tested. That demolished most of my reservations of unit-testing and development agility.
* Meeting some of fine Yahoo US folk during that time: Ryan Kennedy and Dav Glass.
* The internal hackday win led me to meeting David Filo to give him a quick demo. Not often I meet people who are also Wikipedia entries.
