Great question, and the forum poster is likely correct — here's the reasoning:
Dynamic range = maximum undistorted output level − noise floor
So DR can improve if either the noise floor drops, the max output rises, or both.
Why reducing gain can lower the noise floor: The noise a preamp generates is largely set by its input circuitry — the first transistor/op-amp stage, input resistors, etc. That noise is relatively fixed regardless of gain setting. If the gain reduction is implemented after the primary noise-generating stage (e.g., in a later gain stage or output attenuator), the fixed input noise gets amplified less, so the noise floor at the output drops.
Why the maximum output may not drop equally: The output voltage ceiling is often limited by the output buffer/driver stage (which is constrained by the supply rails), not the main gain stage. So if you reduce the gain of an earlier stage, the noise floor falls but the output stage can still swing the same maximum voltage. Result: dynamic range improves by roughly the amount of gain reduction.
A secondary benefit: volume control position. With a lower-gain preamp setting, you typically need to run your volume control higher to reach the same listening level. If the volume pot sits before the main gain stage, a more-open pot means a stronger signal entering that stage relative to its own noise — improving SNR.
