White noise is not a valid stimulus to assess gap perception.

Evidence

As we observe in the top panels of Figure 5–4, when a gap is introduced into a white noise, spectral splatter is negligible, as white noise already has a wide frequency range. The two spectra contain the same frequency components. Although the two noise spectra are very slightly different from each other, using a random sample of white noise each time a stimulus is presented (with or without the gap) prevents the listener from using slight spectral differences to detect the presence of the gap. We can do this because every sample of white noise is different with unique amplitude fluctuations. — Jennifer Lentz, Psychoacoustics- Perception of Normal…, 956