Scale (as described on page 67 of the System Toolkit and page 52 of Condensed) is a rather useful tool for modelling certain kinds of discrepancies of power between entities in a campaign. It seems especially convenient for modelling suddenly-encountered situational differences, like 'the tank you just hijacked is bigger than other vehicles; gain 2 Scale for while using it against enemy soldiers and small vehicles'.
However, there are cases where Scale seems appropriate as a more long-term thing, and one that is not the same either between PCs (e.g. some PCs are fair folk or supers, others are plucky mere mortals) or their assets that are built using the Fractal (e.g. all PCs are starship pilots, and each player designs both a character and a ship).
I'm aware of some game lines that have implementations of taking different Scale for different characters/ships/etc., but they seem to be very ad hoc and often messy (e.g. Mindjammer), and their implementations of effects of scale are not even necessarily in line with the System Toolkit/Condensed mechanical elegance. They also seem rather opaque in terms of what are those modifications of picking Scale during character/vehicle creation are based on.
Well, I'd like to know more about said basis. That is, I want to figure out the general case that can then be modified rather than chase individual narrow examples that have already been modified. And a major part of the general, at least to me, seems to be the (at least approximate) value that should be assigned to Scale in character (or vehicle/asset/etc.) creation.
E.g. I've seen a campaign where Scale (explained as a ladder of degrees of godhood) was worth one Refresh per level. But is that an accurate/fair assessment in terms of actual usefulness of Scale? If not, what is a fairer one? Is there either a RAW precedent for establishing such a number, of a community consensus based on actual play experience across multiple gaming groups and different campaigns?