The New Scientist mag has just published an article about my recent obsession: the flyby anomalies. I've been working hard to explain these anomalies for a couple of years now, and have published a couple of papers that use a new theory I have been developing (based on an earlier suggestion by Milgrom (1999) that inertia may collapse at very low accelerations). My development of inertial MoND involves a far more gradual loss of inertia for low accelerations and is called 'Modified inertia by a Hubble-scale Casimir effect', or MiHsC. It is indeed 'Physics from the edge of the observable universe' and went through the painful but useful process of peer review, which, as usual, drove me crazy. MiHsC predicts the flyby anomalies without the need for adjustable parameters, it is easily testable, and although it violates the equivalence principle (EP) for very low accelerations (10^-10 ms^-2) I'm not worried about that, since the EP is an assumption and hasn't been tested in that regime.
MiHsC isn't perfect though, and it's important that I state that it only seems to work for unbound trajectories, unless all our masses are wrong, and this keeps me up at night, but the point is that I put it through full peer-review and the New Scientist article mentions a load of articles from the arxiv that have not been through peer review, and the only attempted explanation for the flybys so far that has been through peer review (ie: mine) was ignored completely! If interested, see: http://arxiv.org/abs/0806.4159
