I'm sympathetic to the purpose -- I'm less protective of privacy than most here.
My objection is to the means, namely, giving the page's author essentially unlimited access to compute resources on the browser and hoping the author will choose to be moderate.
Until 2008 I browsed the web on a slow machine (Celeron based on a PIII core, probably made in 2000 since the BIOS copyright notice referred to the year 2000) and blocking GA on that machine very drastically reduced the amount of time the browser was unresponsive. So much so as to suggest that even on modern fast hardware, there will probably be some improvement in responsiveness even if the difference will not be immediately noticeable to the user like it was on my old machine.
My objection is to the means, namely, giving the page's author essentially unlimited access to compute resources on the browser and hoping the author will choose to be moderate.
Until 2008 I browsed the web on a slow machine (Celeron based on a PIII core, probably made in 2000 since the BIOS copyright notice referred to the year 2000) and blocking GA on that machine very drastically reduced the amount of time the browser was unresponsive. So much so as to suggest that even on modern fast hardware, there will probably be some improvement in responsiveness even if the difference will not be immediately noticeable to the user like it was on my old machine.