Many possible causes; probably working in concern with each other.
Theorem 1
One oil seals up the combustion process "better" (I almost detest that word because it's difficult to define in a short sentence). That improved cylinder piston/ring/wall sealing ability makes for less contribution to the soot loading, contrasted to the other product, by making for a more efficient combustion process. Soot is the byproduct of any hydrocarbon fuel that is not completely burned. A more efficient process burns fuel more completely, and therefore produces less soot.
Theorem 2
One oil has more detergents than the other, meaning that it cleans "better". (Again - that nagging word). The thought here is that it can scour up what is left over from a previous load quicker than the competing product.
Theorem 3
The disperent packages are different. But know that the oil additive can only absorb what is being produced by the system as a whole. The dispersent oil additive cannot affect the rate of soot production; it is reactive and only can affect the rate of absortion. The more dispersents, the more soot it can hold in suspension before the oil is saturated and the soot falls out (and then comes into the baliwick of the detergents to scrub/scour it from surfaces).
The difference between #2 and #3 is that detergents are cleaners, while dispersents are anti-agglomerates. While they work in concert with each other, they are not techincally doing the same thing. Detergents cannot keep soot from combining. Dispersants cannot clean soot/insolubles from surfaces. The thinking could be broken down into almost a revolving relationship.
Imagine this:
Totaly pure clean, new engine.
Start using engine; fuel is consumed and burned, some soot is produced.
Dispersents try to keep soot in suspension; most of it works.
Some soot falls out; detergents go to work, trying to put soot back into suspension by making it available to the dispersents.
Eventually, one or the other (either the dispersents or the detergents) become overwhelmed, and the cycle is broken, and soot accumulation become gross.
But we cannot avert the concept of soot size, particle wise. It is much more desirable to have 10 particles of soot that are 1um in size (essentially harmless) than to have 1 particle of soot that has agglomerated to 10um in size (certainly damaging to some engine clearances/parts).
I suspect the experiences you are seeing are some combination of the three above, and will be very difficult to ascertain which is the true contributor(s).
I guess that for me, at least, the real questions are this:
How are the wear metals in a used oil analysis?
How is the soot concentration in a used oil analysis?
Until you know these things, it's hard to tell if one product is doing a "better" job than other. (Darn it; there's that word again ...)
