Lisp was a competitive advantage for the company that made him rich in the mid-90s. He wouldn't claim it is for a new company starting out now. He later wrote that Python and Ruby were a competitive advantage. They wouldn't be now. But other languages/platforms are. His claim wasn't that Lisp is and will always be a competitive advantage, but that at any time, there is a powerful platform available that few developers/companies have embraced yet, but that could be a competitive advantage to a company that does.

As for Microsoft, it still is dead in the sense that he articulated in that essay; it doesn't offer any novel/powerful/ubiquitous platforms that new companies could or should build new products/companies on. It barely exists as a mobile platform, nobody has built significant new companies out of Windows desktop software in years, and its cloud platform is just one of several, and far from the biggest/best. He never said it would go out of business, just that it would become a generic BigCo, which is exactly what it is.

He also built a new type of investment company that invested in Airbnb, Dropbox, Twitch, Stripe and many other new startups, before any other VC would look at them, and at a tiny fraction of the valuation that other VCs would invest at. That investment model he pioneered has since been imitated countless times around the world - unsuccessfully, in most cases, because he did such a good job that others couldn't seriously compete.

With all that said, I think there's plenty he's been wrong about, but he's been right about the most important stuff, and deserves admiration for that.