The pontil rod is attached after removal from the mold. This channel is part of the mold (as to "why" I do not know) so the channel does not run through the pontil - the pontil sits atop the channel.
Attached is a photo of a GIV-17 I purchased on Monday with the same channel running across the base. Removal of the pontil rod was often the reason for breakage - might the channel be a means of reducing the amount of surface area to which the rod was attached to the base? Not sure I am explaining this well but the reason that it appears that the channel runs "through" the channel is that this area is recessed and thus the scar is not visible here. With a smaller attachment point perhaps the glassblowers felt it was easier to remove the flask from the rod. Similar functions were utilized in earlier times through the use of a molette....
I think it is a modification in the mold to remove an uneven base left behind by the original molds seam. Maybe the mold became loose on the hinge and a gap opened up, who knows? Mike, (earlyglass), has another flask in a different mold with the same channel in the base while he also has the same mold without the channel... I was thinking they possibly used a tool after coming out of the mold, but I see as pointed out by others that wouldn't make a lot of sense...