Ummm...
I didn't really like his assessment of VisiCalc. It seems to me that ten years or so of development in Excel is bound to blow away the original. Hardly a fair assessment.
By that token, how does Mosaic get on the list? FF or IE should be good candidates there. Seems like a bit of a double standard. About Mosaic he says, "In other words, great software that opened the floodgates." Well, how does that not apply to VisiCalc?
And while UNIX (or Mac OS) may be great, it's still Windows that made computing accessible to the masses. That doesn't count?
Java? Huh? WTF? .NET is Java done right. How the heck does Java get in there?
Why not cite an example of an FFT algorithm? DSP programming drives all the really hardcore scientific applications.
He doesn't seem consistent to me. Excel wins on quality, but Java wins on innovation. Apples and oranges. Either we're talking about great quality or greatly innovative, but he's mixing them.
I dunno... Maybe I'm just getting a bit jaded with all the "Top 10" lists out there.