I agree completely. That's why I opted out. Initially I thought the data was compared directly against competitors but the fact that it is against sites of comparable traffic makes it even less likely that I will sign up.
It seems to me that unless they compare my site's traffic against direct competitors this offers me no benefits and even if they ever were to offer that option it could contradict their search results ranking – ie. why wouldn't my site rank beside my competitor's in organic search results if Google does recognise they are in the same market and offering the same services as me.
One final note; I can't remember their wording for the sign up option, but I do remember it struck me as slightly agressive – ie. sign up or all your hair will fall out and your cat will contract FLV.
Regarding your last note, this is the dialog:
<<Google Analytics Data Sharing Settings
In order to improve your experience with Google products, Google Analytics is updating its data sharing policy. You now have the ability to share your Analytics data with other Google services. This will improve integration, enable additional features in Google's advertising services (including Google Analytics, AdWords and AdSense) and improve your experience with these products.
Press "Accept" to enable data sharing between Google Analytics and other Google services or for additional options, choose "More data sharing options".
Remind me later | More data sharing options | Do not share data | Learn more>>
Hmm.. I seem to remember there was definitely a note regarding 'If you do not share etc.'
Maybe it was something on this page...
it's more useful to Google than Google Analytics users – by making the data public Google has the opportunity to share granular and raw data within Google and across departments without violating privacy or TOS
The benchmark data's usefullness depends on your point of view, ofcourse.
Personally, i found i very nice to see that 1 of the sites i've build (oranjebud.nl) got around +1.018,56% more visitors then the benchmark (10 times more).
That is 'nice-to-know' but more usefull was data about another site (tomwaakopreijers.nl), which underperformed, according to the benchmark. This motivates me to look deeper into the 'normal' Analytics data and improve the site accordingly.
Concluding, i see the benchmark data as a nice addition to an already very powerfull tool and hope to see improvements in time.
I tried two hotel websites I develop and the bench marks were totally different. The one which gets a high amount of hits had a different benchmark and the site which got lower hits. Actually this did not make sense.