The band analyser works now, in fact remarkably well - at least on the dumps we have. It is very good in identifying the correct flux transition periods used, even if a normal average calculation yields a very different result. There is still room for improvement, perhaps a slightly more scientific approach to find the distribution and correct trend analysis, but it works well enough for now.
Taking a look at the statistics from the protection track of Dungeon Master reveals that it is indeed using flux periods that could belong to different bit patterns.
Still to do on this part is find the bands that are related to each other, basically being more or less multiples of each other. Finding their common base gives the cell size that was used.