Yea. It's way more complicated than temperature. The highest ratios seem to be from shallow clouds with a cloud-base dewpoint around 10-15 degrees. When dewpoints at the cloud base are in the single digits or lower the snow gets more dense again as you have mostly diamond dust plates. It's lighter than a wet snow, but denser than fluff. Lots of drifting decreases the ratio as well because the aggregate flakes get broken up into individual crystals and land flat. Snow that has blown off the roof can be especially dense even when it's cold. Wind-sculpted cornices can be rock-hard and exceptionally dense, probably because the loose flakes all blow away leaving very dense semi-fused crystals behind.
Also, snow that forms in really deep strong updrafts is denser due to rimed flakes. Light-intensity lake-effect is more fluffy than the stuff that falls from really intense bands. It seems a bit of a nonlinear curve. Very weak vertical motion will produce mostly plates due to supersaturation not being great enough for dendrites. Moderate vertical motion will produce big dendrite flakes that form aggregates and really boost the fluff factor. Extreme vertical motion leads to rimed flakes and eventually graupel, which is dense.