You can use named nets and an iterated instance to do
this efficiently. Each "pin" (consider it a pin-stack) has an
ordered list. So, like, a ripple counter could have one
iterated DFF instance, the D pin (stack) might have
<Dprime,Q<0:6> while the output rank is Q<0:7> and
the CK pin is a single shorting wire to the clock source.
Much better than a wireball, as long as you keep the
bookkeeping straight.
Probing that can be a bitch, though, so I usually will take
those bundled nets out to singlets with a trivial load element
that lets me plot-by-pick, somewhere that doesn't mess
with core schematic prettiness.