It's a question of audiences.
To an expert on bitcoin, the description of "the address with the most Bitcoins" is self-explanatory and, yes, has one well-defined answer that it can be determined from the block-chain alone. If you're this person, I can understand why you think I'm being pedantic, and I'm sorry if I came off that way.
But someone who says "If anybody tells you Bitcoin is anonymous, they haven't the first clue how Bitcoin works" isn't talking to an expert on bitcoin. They're talking to someone that they fear may not know enough to avoid being easily swayed by the kind of inexact language that a non-expert doesn't even know that they're looking for. They're talking to the exact kind of audience that doesn't know that "person" and "address" are not interchangable
This person may then go on to claim, based on this evidence, that (1) no anonymity can ever be obtained by bitcoin (which isn't true in theory or in practise for the reasons I mention) and (2) that we know "who" (as in person, not address) has all of the bitcoins. But that's just not true, and it's a really, really important detail to someone that makes policy decisions, or that wants to decide if or how to use the technology.
It's the same problem with equating an IP address with a human. You may use exact language when you talk about these things, and you may understand that when someone offhandedly points at an IP address on a diagram and says "...but this guy over here..." that "guy" doesn't refer to an actual human being. But you aren't any of the grandparents to this conversation, nor are you the hypothetical uneducated observer (to whom I mean no ill will; we all are, given the right conversation). Most importantly, you're not ninetax, who is clearly referring to a person and not to an address, nor are you nknighthb, who is clearly referring to the notion of an address rather than a person.