This Letter revises the recent claims of a dynamical constraint on massive dark clusters from the Galactic globular cluster luminosity function (GCLF). We point out that this argument is invalid for two reasons. The first point is that with a proper sample of globular clusters, the GCLF presents a significant trend with Galactocentric position, and second, the disruption timescale of globular clusters by encounters with these massive objects depends strongly on many uncertain parameters. For the usual halo model of Cadwell & Ostriker, it may be smaller by only a factor of 2 for distant globular clusters. Therefore, the existence of massive dark clusters in the halo could be possible. This provides a scenario in which the formation of dark clusters and globular clusters may have a similar origin.