Second roll out test

Managed to find a spot that was suitable for a roll out test.  It is a few km away but one of the better smooth stretches of pavement on the island.  A pull out on one side the of road for turning around  and a T intersection an couple hundred meters from the test section to turn around and come back.   Only issue is the road has a slight curve in it but that didn't seem to matter.  The test section I used was about 100 meters long with a slight downhill to start then a tiny bit of flat and then gradually uphill.  I would reach 12-13 kph coasting down then roll to a stop.  Testing equipment consisted of three sticks about a half meter long.  At the top of the hill I laid one of these sticks in the bushes on the side of the road.  To start the test I would line up so I could sight along the stick and then release the brake.  The second stick was  used to mark how far I rolled out.  The third stick was to mark any variation from the first roll out.  I did three roll outs with the setting I arrived with.  All were within about a half meter.  I reset the toe in by one turn and did it again.  The difference was a half a velomobile length from the previous test.  I did this a couple of more times just to be sure and it was always the same.  The test procedure was very repeatable.  The end result was that I did not have front wheel alignment set optimally.  After a few more trials I narrowed the setting down to two possibilities.  While they were close I always rolled further with the one setting by more than the margin of error.  The real question is whether this effort was worth it.  I very much suspect so just from the fact your tires won't wear out as quickly if the setting was off.  Even just using the difference in distance rolled this would amount to at least 10 meters per km.  It is most likely considerably more when you apply effort or coast down hills(go faster).  Certainly enough to be noted by a riding buddy if they are slower than you or you are having trouble keeping up. A small improvement in the test procedure would be a better way of measuring the start position.  I will try a round tube taped of the side of the velo which I sight down to a line on the road.  This should get the start point to within a cm or less.