The problem with the term "labialized velar" is that it is used
rather vaguely to cover three rather different things:
a) Velars articulated with rounded lips: place the lips as if
for whistling, then pronounce the velar stop.
b) Actual labialized velars: bring the centers of the lips
together almost as if to articulate an _m_, but with a small
opening, then articulate the velar stop.
c) Co-articulated labial-velar stops: first close the lips as if
to articulate _p_ or _b_, then while maintaining this closure place
the tongue in the position for articulating _k_ or_g_, then release
both closures simultaneously while emitting air.
In my experience only (c) is actually difficult to learn.
Older phoneticians, as well as many modern phonologists and
certainly Indo-European philologists past and present were
not well aware of the distinction between (a) and (b) and
in general not aware of (c) at all, in spite of the fact
that the developments in Greek and Celtic imply that the
Proto-Indo-European "labiovelars" were of type (c). The
same is alas no doubt true of Tolkien as well.
P.S. In Latin and Common Germanic the PIE "labiovelars"
had probably developed into clusters with [w]. It should
be noted that a _phonetic_ cluster can still function
_phonemically_ as a unit.