A note on global suprema of band-limited spherical random functions