I would say that it isn't a big deal if others have done very related work before, especially if their data don't support their conclusions. The PhD doesn't necessarily have to be a novel contribution to the field; however, it should be significant. Depending on how you approach the research, re-interpreting new data in a manner more consistent with them could certainly be significant. Plus, if the data work out the way you think they will, you should be able to imagine the next course of action to take, either on your own or in discussion with lab members and your supervisor.
Discuss this issue, and the other problems you mentioned, with your supervisor. Generally speaking, your success is his success, and he should be keenly aware of this already.