Correct me if I'm wrong, but the paper published three years ago unequivocally states that CAT incidents have increased and are likely to continue do so based on data gathered. The mesh grid used is too coarse to provide specific predictions for individual aircraft, but that is to be expected.
Given that the gross details are already mapped and conclusions drawn, it seems that the best way forward would be to equip as many aircraft as possible transiting these routes with calibrated accelerometers which broadcast the data back to collection point(s). That way you could get a precise GPS position and altitude linked to an objective measurement of the turbulence experienced.
Such data could then be overlaid on the existing mesh calculations to verify their accuracy and improve predictions. New sources of data are needed rather than simply rehashing existing data sets at finer scales.