I realize it has become fashionable on the boards to look about for anyone that you can "hey noob", but I am not exactly new to this game. I've made probably 40 fortresses just in the 3D era, and probably that many in 2D. I am fully aware that magma cools into obsidian, and that cooling can occur prior to embark (ie the obsidian caps on glacial magma pipes).
This is the first time I have ever encountered this particular phenomenon. Much like the other posters I think this may be a bug, and thus should be brought to Toady's attention.
If he says "Yeah, it's intended" then its a feature. I'll then post a thread in suggestions to amend either magma pool depth or site finder reporting.
So until we actually hear from him, let's keep the thread clean of non-constructive conjecture...m'kay?
edit: I regenerated the same world to tinker a bit. While poking around in the same spot, I noticed that there was not even a z-level of obsidian, there was just a layer of "obsidian cavern floor": the underlying rock was alunite. If obsidian had cooled, there would be mineable obsidian.
It is as if the game generated the very bottom of a magma pool (its lowest open level + floor) but did not fill it, make an obsidian shell, or extend it any higher. To use Keizo's terminology, it was a 0-level magma pool.
The nearby magma pipe is normal, as are the nearest 4 other magma pools.
The only oddity may be that it appears on one level that pool might straddle a biome boundary.