I'm having a hard time parsing this, and should probably clarify my comment based on what I've seen:

802.11 base spec states the need for a beacon frame to be sent from the access points for various reasons. Part of this beacon frame is the SSID.

Some vendors decided to allow that SSID to be set to NULL instead of the actual network name, thus becoming the "hidden SSID" option so many access points have. This turns off one, of I believe five different frames that have the SSID in them. Problem is when this is turned off on the access point, clients tend to broadcast the SSID in the clear to try and find the access point. Buggy behavior in this situation with Windows XP (pre WPA2 patch) led to the "Free Public Wifi" propagation issue (also tied to a bug where Windows would create an ad-hoc network on it's own when it couldn't find the base station).

802.11i further clarified how the beacon frame needs to be handled. This included the need for the SSID in the beacon frame to match the SSID in other types, as per the handshaking that happens to make WPA2 secure. This comes into play mostly in multiple access point environments, to allow clients to roam securely without a full disconnect and reconnect during every roam.

The "may" choose to refuse to communication statement seems to be in there as a warning that equipment following the spec fully may programmatically choose to not communicate, due to mismatching SSID info being sent in different frames (NULL vs the actual name).

802.11n had some additional clarifying language added.

I should probably get back to work though instead of looking at 802.11 tech spec documents. They are very dry and hard to parse at times. And definitely wouldn't fit into 140 characters to win some twitter fight.