Oh saved another one.  I tried a number of different ways to handle the input before resting on this as the best.  There are a number interesting ways to build the molecule without rotating as this does but this seemed to work best for me, everything else I was consistently off by one.  Limiting factor in terms of reducing it further with this setup is getting the last arm out of the way of the molecule, which even if I solved that it causes accidental bonds.  MANY of the other ways I tried to set this up ended up being off by 1 in terms of being able to get the molecule on the triplex but not having an in creation molecule hit the triplex by mistake.  I am guessing you can somehow get lower, but not fully certain how based on the above limiting factors.