Lot of people on Twitter are currently wondering why you can still access Wikipedia by disabling Javascript. It's not that we're cheap, or want a backdoor. There's many factors at play here.
The way Wikipedia and Wikimedia sites operation with very little hardware is through intense caching. Almost all anonymous requests to Wikipedia hit one of our many squid servers, never reaching Apache.
Doing a full blackout would pollute this cache (Meaning that this blackout page would stick around for days, even weeks after), and may not even show up if the page is already cached.
During these last few days the operations team at Wikimedia has been in constant contact with companies such as Google to make sure that this event does not affect out search rankings. There are concerns that 503ing pages would affect rankings.
johnduhart is not a WMF employee, just a volunteer
@johnduhart Thanks for this. I'd have thought adding an ACL to the Squid config would have made returning pretty much anything relatively easy. I'm no expert (far from it), but I can't imagine that ACL decisions affect the cache contents?
Search performance is probably the more important concern, but I also saw advice posted by Google staff on their support forums stating that 503s are the right thing to do (with provisos about robots.txt and such).
In any case, better this blackout than none at all. Thanks for the information.