By now most of us will have seen the joint ESRI/Google announcements at Where 2.0 about the upcoming ability to use Google Earth and Google Maps to discover, locate and portray geo-data hosted in ESRI ver 9.3 geo-server stacks to be released later this year.
In the interest of balance (and with no desire to provoke any flame wars) I draw your attention to another announcement at Where 2.0 addressing Geoserver's upcoming 1.7.0 release with similar capability - see http://blog.geoserver.org/2008/05/13/geoserver-and-googles-geo-search/
The point here is not about commercial versus free versus proprietary versus open source software. It's about two other more important things.
First: the maturation and acceptance of standards. Google has embraced the OGC specifications process and KML is now an anointed OGC standard; ESRI has embraced the OGC specifications standard and is positioning its products appropriately (e.g. ver 9.3 will also provide a fully-compliant web feature service as well as full support for PostgreSQL/PostGIS back-ends). The more this momentum build the more of this joined-up capability we'll all benefit from.
Second: the maturation and acceptance of visualization tools for non-specialists. If the Google/ESRI announcement had happened 5 years ago there'd have been barely a ripple outside the hard-core GIS community. The fact that The Rest of Us can now contemplate being able to publish geo-information (whether social and community based, conservation-related or whatever) --and-- have it discovered, integrated, employed and appreciated by non-specialists audiences via attractive spinny-globe-type-products is a powerful combination.
What is crucial here, and to me will make or break this effort, is the ready availability of easy-to-use tools to take my existing geo-data and get it into a search-able, discover-able, use-able form without having to have a masters degree and without having to learn to speak techno-babble. And that is equally true whether 'I' am individual Mick Wilson or whether 'I' am a multi-national conglomerate.
We need to keep lowering the bar to the publishing of real-world data about real-world events by real-world-people.