
The Wireshark (nee Ethereal) 'revelation' dragged me here (too), as the first two clangers weren't sufficient to get me motivated enough to write an ignorable comment about the show.
MS-Office 'not that expensive' and the whole office thing 'not important' -- this seems wrong, once you've scaled the numbers to match a reasonable-sized organisation. And the bit about needing only 5% of your documents being 'dubious' with a different office application being an imperative to stick with MS-Office, I guess that's down to perspective. From the same starting position you could argue that this is the primary motivator to shift away from a proprietary document format.
Segue to the brief mention of Sharepoint - my take on that shift was partly SAAS, partly to further limit the potential impact of the OpenOffice and GoogleApps of the world. Sharepoint's a one-way street as far as organisation document management goes - really, there's no coming back from that decision - and from what I've heard the document format for storage is the not-yet-ratified OOXML format. Unsettling stuff.
HP 'make loads of money from HPUX' .. this claim is simply surreal. PA-RISC is dead, HP-UX releases are consistently delayed (the most recent was 12 months late, IIRC), Intel's commitment for Itanium is ever doubtful, and just look at HP's annual reports on revenue breakdown to see how important this business is to the printer company (and by extension, to Intel). None of this is reassuring for existing users, and utterly not enticing for potential new customers.
As to clouds .. hmmm .. maybe we could have centralised large-scale servers with everyone attached via long bits of wire .. we could call those servers mainframes, and the client equipment could be some kind of dumb terminal.
Oh, and you need a flannel-slippers icon.