As well as being a 'traffic awareness' device, FLARM is also a flight recorder. If the concerns expressed above were a real, rather than theoretical problem, then discontinuities in the recorded track would be evident. A close up view of a recent trace, with turns at 45 to 55 degrees bank in both directions
shows no such features, and I have never seen worrying discontinuities in many other records.
As far a white aircraft being 'madness', there is a vast difference between a white glider being 'invisible' and the colour being less than optimal. Many aircraft of non-composite construction are primarily white in colour. I have no difficulty when flying gliders in spotting other ones (possibly because I am looking for them as one aid the using the energy of the air), and apply similar techniques when power flying. In my experience, the shape of conflicting traffic is visible before its colour can be distinguished. Some colours have been shown to be optimal for being more easily seen at closer ranges, but the limited evidence (and more research would be very useful) is that colour has very little effect on the maximum range of detection.
One could argue more cogently that it is 'madness' to perform aerobatics in an area not under full traffic control, since the Lookout part of the HASSELL check is only useful for the traffic in visual range, and a good lookout is unlikely to be maintained during the manouvring period. The saving grace is that an aircraft performing aerobatics is much easier to see than one maintaining a constant heading (which also applies to gliders which in general seldom fly on a steady heading for long) and conflicting traffic more likely to avoid an airprox situation.