Distant Voice,
I really hope you are just ignoring my point, not missing it. I thought it was simple enough. I didn't say they made up figures or any such thing. I simply pointed out that the percentages are meaningles without understanding the underlying data. And that making a comparison between Tornado (flying at low level, not under radar control, doing simulated or actual attacks, etc) and an airliner flying under radar control in controlled airspace is hardly a reasonable (or statistically sound) comparison of percentages.
Similarly, saying that Tornado accounts for X percent of airprox reports and Harrier only Y is only meaningful if you understand what their comparible flying hours are. If you really need me to explain, it there were twice as many Tornado hours it would not be surprising is they had twice as many airprox incidents.
Get it?