Sometimes the argument on accuracy of the machines can become a little meaningless.
At the end of the day its about how much good tissue was spared from the effects of radiation, while giving the tumour the required dose.
Machine Accuracy is only one part of the puzzle when it comes to sparing good tissue.
Software versions, Power of the beam, CT Scan, and MRI accuracy, Merging CT and MRI's, patient positioning, size and shape of the beam, size and shape of the tumour, Patient movement, head frame flexing, movement of organs, radiosurgeons preferences on treatment, experience of the radiosurgical team... even your brain moves in your skull..
So when comparing the machines, to look at mechanical accuracy is a little simplistic.
The radiosurgeons goals is to kill the tumour and to reduce damage to good tissue as much as possible.
This is why I quoted the paper that I did. It was current Feb 2012, and demonstrated that Cyberknife specifically for acoustic neuromas appears to have a significant advantage over Brainlab/Novalis for sparing good tissue.
http://www.ncbi.nlm.nih.gov/pubmed/21892741The Novalis/Brainlabs ability to modify the beam shape to match the tumour shape, does have advantages.
And if you compare Cyberknife to Novalis/Brainlab on a large odd shaped, aggressively growing tumour, the result could be different. One machine is not necessarily better than the other for all tumours.
I hope that goes some way to explaining why there is conflicting information about the machines.
Nobody should ever take advice on public forum as being accurate.
Healthy debate is good.
It will encourage people to investigate further, so they can come to their own conclusions and treatment choice.