GPU version of Gravitational Wave search

I think this has come up before and a good reason for why it won't happen has been given.

But ... Christophe Choquet seems like a smart cookie, he did a nice job on the FGRP GPU app. Couldn't you throw the GW problem at him and see what happens?

Maybe frame it like an impossible problem that everyone has failed to solve. Surely not even he can do it ... ?? :)