...and who says e-mail spam filtering works?

The following excerpt comes from MSNBC's "The Red Tape Chronicles" :

[Let me begin by saying that you cannot make this stuff up!]

Friday: 10 Oct 2008
(http://redtape.msnbc.com/2008/10/att-customer-ca.html#posts)

AT&T reserves the right to change its terms of service by sending its Internet service customers an e-mail. Apparently, it also reserves the right to deposit those e-mails into its customers’ junk mail folders.

Last month, AT&T made some controversial changes to its Internet policies. Verbiage indicating that high-bandwidth users might experience some intentional slowdowns irritated some techies; another section that forces customers to use binding arbitration to resolve disputes annoyed consumer organizations; and an L.A. Times reporter bristled at the size of the full new agreement -- 2,500 pages.

But Lance Mead, an AT&T Internet customer from Encino, Calif., almost missed the entire controversy. His notification of the new terms of service was sent via e-mail on Sept. 18, but AT&T's own spam filters trapped the e-mail as spam and deposited it in his junk mail folder, he said. On a whim, he checked the folder and spotted the notice. He was furious.

Someone -- anyone -- please tell me how this is not proof positive the entire premise behind e-mail spam filtering is seriously flawed? I completely understand that mistakes happen. However, these "mistakes" are also considered "false positives." In the "e-mail game" it is the false-positives that cost business real money. Is it really the end of the world if 5% to 10% of the e-mail received in your inbox is spam? Probably not. It is unnecessary, annoying, and unproductive to be forced to wade through spam, but missing an important e-mail thanks to the flawed concept of filters, a.k.a. guessing machines, should be considered absolutely unacceptable.