Interesting stats those. They don't seem to mesh at all with the ones on page 4, which is showing around 100 accidents in 2005 (with a downward trend), though with around 12 fatalities in 10 fatal accidents (very hard to tell - wild XXXX guess).
So assuming you are talking about fatal accidents only (again, the preceeding slides don't make it clear) on page 7, and you add all the GA sub-types up then they come out worse than the ultralights, but around the same for all sport types (+gyros +gliders).
But then if you take the fatal accidents/100000hrs for both types then ultralights come out around (ballpark, can't really tell - give numbers damn it) all of GA taken together.
In conclusion I think that presentation is an excellent example of comparing apples with elephants. I think the fatal accidents/100,000 hours is probably the best point of comparision (even though its the worst for us), though even that is a tad dubious.
I still think statistically insignificant difference is a hard call to beat. Why is there no smiley-with-head-stuck-in-sand emoticon?