Well, if the Tertiary no longer exists (side note: given that these divisions are all purely arbitrary, an attempt to create discrete categories where none exist, why do they need to keep changing them?) why still use it for the name of the K/T boundary?
Inertia. K/T rather flows of the tongue, everyone knows what it refers to, and it's unambiguous (it's not like there
is a Cretaceous-Triassic boundary!), so people see little reason to stop using it. Still, K/Pg is making inroads, at least in the professional press.
I would dispute that the divisions of geological time are
purely arbitrary. They were originally defined with respect to the appearance and disappearance of variou fossil groups, and thus their boundaries tend to fall at evolutionary turnpoints. The K/T is a case in point - the Cretaceous period gives way to the Palaeogene (and the Mesozoic era to the Cenozoic) at a mass extinction that changed the course of life forever.
As for the Tertiary, it was originally part of a very simple division of strata into Primary (roughly = Precambrian), Secondary (=Palaeozoic and Mesozoic), Tertiary, and Quartenary (=Pleistocene and Holocene, the last ~1.65Ma). Forever reason, the last two survived into more modern schemes as the grossly unequal periods of the Cenozoic era.
However, the basic reason for having a geological timescale at all is pragmatic, and it's not very practical to break down a larger unit so unequally - the Tertiary was close enough to synonymous with the Cenozoic, whereas the Quarternary had a length hugely smaller than any other period. It was found more useful to divide the Cenozoic into the Palaeogene (65-23 Mya) and Neogene (23-0 Mya) periods, but then the Tertiary and Quarternary had to go to make room in the hierarchy.
Exit Tertiary.