@all web sites affected?
I don't know, but given that the whole *purpose* of the company concerned was to identify individuals based on the IP address, then ruling that IP addresses can be considered personal data is reasonable.
But as already pointed out, in most cases all it identifies is the home NAT router, not which machines are behind it (wired or wifi), and more importantly, not who is actually using them.
That was one of the rather nasty aspects of the recent "Digital Economy Act" and elsewhere, to make the subscriber liable for abuse of their internet connection. Something that may initially appear to be reasonable, but one that on closer inspection is not. For example, who is responsible for:
(1) OS holes and similar that lead to a machine becoming a botnet, and abusing the connection?
(2) AV software that fails to identify malware leading to (1)?
(3) Security protocol flaws that lead to ease of abuse of wifi (e.g. original WEP), who then pays for replacing otherwise good equipment that can no longer be considered "secure"?
(4) Actions of children. After all, if the parents of child killers don't face gaol for their child's actions, why should they for something as trivial as copyright infringement?