Well it's hardly a full-scale scientific trial, but someone is actually putting this to the test right now: there is a Dan McLaughlin who has set out to do 10,000 hours of deliberate golf practice from complete novice... and we'll just have to wait and see how good they end up. So far they've done around 3,300 hours.
The Freakonomics people have some commentary on his programme here and some writing on Ericcson's theories here. Reading the Ericsson paper they cite, the only mention of 10,000 hours seems to be
In the following discussion we reexamine the evidence ... for the role
of innate differences in the attainment of expert performance. The
dichotomy between characteristics that can be modified and those that
cannot may not be valid when we examine the effects of over 10,000 h
of deliberate practice extended over more than a decade.
and the one cited by @Josh's answer also contains only one mention:
By the age of 20, the best musicians had spent over 10,000 hours
practicing, which averages 2,500 and 5,000 hours more than two
less-accomplished groups of musicians at the same academy,
So really all this 10,000 hours figure is, is a guide to the magnitude of the amount of practicing needed to significantly influence expertise, at least in Ericcsons's studies. However, it's a memorable enough soundbite that it seems to have taken on a life of it's own. Further, once the the "deliberate practice" aspect has been forgotten (e.g see comments on original post!) people find it useful as a tool to justify complacency ("well I've been doing this job for a few years and at least 10000 hours easily; so of course I must be an expert now").