VA-Spec functional impact + pathogenicity annotatoins are 'available' on score sets without calibrations

[bencap]

A previous change to API logic sent back empty lists rather than null values for score sets with no calibrations. Because empty lists are truthy in JS, this had the unintended effect of making UI code which checked for calibrations believe that score sets with empty calibrations lists had calibrations.

We should update conditional checks on calibrations to be based on list length rather than existence.