I wasn't sure about the benefits of MOVE TO until I did a drop constraint and then inspected the table.

The data had shifted onto the desired filegroup so I would hazard a guess that the subsequent creation of the clustered index is dedicated to creating the key and not shifting data at the same time.

Before anyone mentions it, it has occurred to me that rather than shift data onto filegroups I would have been better off shifting it onto partitions on filegroups. That said it was a quick down and dirty process and it would work on legacy SQL2000 database if required.