Some of the webserver log analysis tools I use, Analog and Visitors among them, show huge, huge discrepancies in numbers culled from the actual logs vs. what Google Analytics shows me. By that token I have concluded that people are already pretty successful at evading GA tracking to a large degree, though probably not to supercookie levels.
If you mean the raw server requests, a lot of that will come from bots/etc without a javascript engine. Blocking GA is probably a growing thing, but it doesn't yet seem to be huge.
Analog tracks the rawest, sure, but Visitors and others, for all their relative flaws, do track crawlers and related. They seem to me to track enough noise that the discrepancy isn't so easily attributable.
