I've been loving coding with MongoDB since I picked it up; I know there are worries, and I come into it with two eyes open, but so far it's been far more useful for my needs than MySQL.
Part of the reason for this is that applications I work on tend to be very document-centric, and where they're relational, they're relational in a way that lends itself very well to NoSQL (membership lists, etc). It's probably not an exaggeration to say that I'm querying the database 10x less than I would be with an Elgg-style entity-metadata-relationships table.
idno is MongoDB-based, including heavy use of GridFS, whereas the latakoo stack is not (I don't think this is the right place to write about latakoo's tech, but I may do on the official blog at some point). However, if my experiences continue to be great, I'll think about doing some more formal testing with the team. With the likes of IBM throwing their weight behind MongoDB (given their recent acquisition of SoftLayer, could 10gen be a target too?), this might be the right time to begin taking it more seriously.