Depends on how you're comparing. Based off the VERY limited data for clo of down fill powers I assumed a linear regression and get about 2.1clo/oz-m^2 for 900fp and only .9 clo/oz-m^2 for 600. That's ~2.33X (233%) the insulating power of 600fp. 900fp has less feather shaft than lower quality fp which may account for clo values increasing at a greater rate than weight savings.
If we assume that 900 and 600fp have same clo, just different weight, then 1 oz of 900fp is 150% as insulating as 1 oz of 600fp, ie 50% warmer for the weight (900/600=1.5). If you want the "same" insulation value, you need ~66% as much 900fp down as 600fp.
For real world applications I would split the difference and come up with 900fp down being approximately twice (2x=200%) the quality of 600fp. Meaning you can use half the weight of 900fp to achieve the same warmth as 600fp
So lets do a cost analysis:
Lets say we're making a down hood (keeping numbers simple just as illustration) and we want 1 inch of loft and our pattern calls for 600 square inches of head coverage (I know this may not be accurate, bear with me it's just to illustrate a point). This gives us a nice 600 cubic inches to fill the hood. I'm going to assume the conservative numbers based solely off weight savings and say that 900fp and 600fp have same insulating value just different densities.
We need 600 cubic inches of down:
600fp=>600c.i./600fp*$3/oz= $3.00 @ 1.0 oz and 1" loft
900fp=>600c.i./900fp*$10/oz= $6.67 @ .67 oz and 1" loft
900fp=>900c.i./900fp*$10/oz= $10.00 @ 1.0 oz and 1.5" loft
So based off performance alone 900fp is at most 2.25 the cost of 900fp. This goes down if you assume that clo/inch increases as down quality increases. You do pay for quality, but that quality comes with the benefit of more warmth and less weight.
EDIT: Rog, you're probably correct about damp conditions and lower quality down maintaining it's loft better due to the extra quill inherent.