1) made gH/gN ratio same in all programs (in dynss and in all reldataX)
2) added decimal precision to gH
3) added decimal precision to d2 in relax
4) renamed dyn9omdc --> dyn_omdc (renamed old dyn_omdc --> dyn0omdc)
5) included in the model selection procedure forced comparison of 
LS_00 with LS_tl and LS_ex --> have to see if the fix covers all possible cases 
(dynmsel_f.m)