Based off stuff I learned from the sources SumnerH has cited, I'd think you need less than half of the hops in a 1.5 gallon batch than you'd need for the 3 gallon batch.
My reasoning is that you'd need less extract or grain to reach the same gravity, and would thus have less break material. Since the break material serves as nucleation sites for the isoalphas and oxidized betas (if I understood Palmer right), that means better hop utilization.
You're talking about doing a 1.5 gallon batch at some gravity (say, 1.040) vs. a 3 gallon batch at the same gravity? Basically, brewing different quantities of total beer. That's not what I understood OP was doing, but to address that scenario:
(long boring response; feel free to skip):
First, it's adsorption, not nucleation--nucleation refers to a thermodynamic phase change. If you're talking about beer, pouring salt into a carbonated beer is a good example--it provides places for a dissolved chemical to undergo a phase change, form a bunch of bubbles, and float away. Adsorption is when molecules of one chemical stick to another; in our case, isoalphas adsorb to break material, which then precipitates out (carrying the isoalphas with it).
With less break material you'll lose less isoalphas, but you have less water so the _concentration_ of break material would be the same in a 1.040 1.5 gallon batch as in a 1.040 3 gallon batch--so though you'd lose less total isoalphas, you'd lose about the same proportion.
In practice with extract, the break material is minimal enough that it's not really worth considering (see BBR's IBU measurements, for instance).
Since you're cutting your volume in half, you'd need at most half the amount of hops. But since you'll have less break material, you can use slightly less than half.
Or am I missing something?
He's only boiling 1.5 gallons but he's still making 3-gallon batch (topping off with water). So he's boiling 50% of the batch size.
Previously he was boiling 3 gallons; he didn't say how big his batch was, but 3-gallons represents a "scaled back" size. Assuming he was brewing typical 5 gallon batches before, then he was boiling 60% of the batch size.
So, in theory, he has slightly more concentrated break material in the boil than he did before, and would need slightly more hops per gallon than before--but going from 50% to 60% of final volume is a very modest change in boil density.
I'd go with the same hops per gallon before (basically, 3/5 of the total hops if he was doing 5 gallon batches and is now doing 3 gallon batchs--if a recipe needed 2 oz of a hop for 5 gallons, I'd go with 1.2 oz for 3 gallons.), personally.