Please read this rewrite of your article as if you were living in, lets say, 1991. Tell me how the this turn out (because NeXT and Steve Jobs did all the things you are asking Apple and Steve Jobs to do now). No other argument could be better than your own words and a god understanding of history.
Rewrite:
Jon Shirley, the only CEO of Microsoft to be universally identified as a grown-up, told me during his last week at the helm that the best thing that had ever happened for Microsoft's language business was Borland International and its language business. Borland, under the wacky Philippe Kahn, gave Microsoft fits. Borland languages were often better than Microsoft's, and always cheaper. The result was that Microsoft, even though it continued to be the larger player in that business, couldn't coast. Borland forced Microsoft to be a better company for its customers, which is exactly what Jon Shirley was talking about. Now look at the problems Microsoft has today, and you'll see that they all come down to a lack of credible competition. Netscape was good and made Internet Explorer better, but today nobody at Microsoft even pays attention to Netscape, just as they no longer pay attention to Novell in networking or, alas, even Borland in languages. Microsoft has killed all the competitors, or at least cowed them to the extent that there is now plenty of excess bandwidth in Redmond for megalomania. This is bad for Microsoft and for its customers. Jon Shirley knew that, but I don't think that Gates or Ballmer do. So the best thing for Microsoft would be a formidable competitor. This kinda sorta exists in Linux, except that Linux isn't organized in any sense, and Linux attacks only Microsoft provinces, not the homeland itself. What is needed is competition for the desktop, and for that, there is really only one other game in town. I think NeXT should market a version of NEXTSTEP for Intel computers.
What I'm talking about a version of NEXTSTEP intended to run on regular PCs. So there would be no NeXT ROMs and no head-to-head competition with NeXT hardware. Most of the required middleware could be bought or licensed by NeXT from Abacus Research and Development Inc. (ARDI) in Albuquerque, New Mexico. In a few months, NeXT could have a perfectly fine Intel port of NEXTSTEP ready for market.
Let me point out here that I have no idea that such an Intel port of NEXTSTEP even exists. I have no inside information. This is just what I would like to happen.
For those who aren't familiar with NEXTSTEP, it is a full implementation of BSD Unix with a Macintosh-like front end, which is to say world class inside and out. NEXTSTEP is faster, smarter, prettier, and easier to use than any version of Windows. In short, it is exactly the competitor Microsoft needs. And the timing couldn't be better. Microsoft's settlement with the Department of Justice, no matter which version finally sticks, will make it easier for competitive operating systems. While Microsoft will still dominate, it will likely not be in its old position of being able to threaten with death hardware companies that take a divergent path. In fact, Microsoft's lawyers at least would love a credible NEXTSTEP for Intel because it would make it appear as though Microsoft actually has competition.
There simply is no technical problem with porting NEXTSTEP to alternate hardware. Where there is a problem is in getting everyone to see what a great idea it really is. Steve Jobs of NeXT has to worry that the new version would hurt NeXT hardware sales. But I don't believe that would be the case. Let's say someone started a sports car company using engines and transmissions bought from Porsche. How many people would buy that new car, the Belchfire 400, over a Porsche with similar power and performance? None. Porsche buyers buy Porsches for the brand as much as for the engine. Belchfire 400 buyers would come from the ranks of Corvette and Viper owners, not Porsche owners. Same with OS X on Intel. Dell, Gateway, and Compaq users are the target market. NeXT users will always buy NeXTstations. This is what Steve Jobs has to come to understand.
The other thing Steve has to accept is that NEXTSTEP on Intel has to be just as modern as any other NEXTSTEP. He'd be tempted to keep the Intel code one version behind to give NeXT a built-in advantage, but that wouldn't be good. It would hobble the product and hobble the marketing at the same time. For NeXT to sell a crappy product on Intel would be like Disney selling porno movies in Thailand. It would change how NeXT defines itself. That would result in a lot of unhappy customers that either NeXT would have to spend a lot of time listening to, or NeXT would have to change its culture to tune out all those customers and become just like Microsoft. Go back to the last paragraph and read it again, Steve. NEXTSTEP on Intel is no threat to NeXT hardware.
And while it might be easy to throw a shrink-wrapped version of NEXTSTEP on store shelves and essentially compete with Linux for the OS conversion market, what is really needed is an OEM strategy. NeXT has to get name brand vendors like Dell, Compaq, and Sony to sell boxes that come with NEXTSTEP already loaded. The Microsoft settlement will allow it, but unless companies actually do it, any advantage over Redmond will be lost. I think Dell, especially, would be in a stronger negotiating position with Microsoft if NEXTSTEP was an option on its price list.
The upside for NeXT is enormous. Suddenly, their software budget is leveraged across a much larger number of units, making the company more profitable and able to spend even more on making the software better. And there is always the prospect that NEXTSTEP will have some real impact on the market, making life harder for Microsoft and making Microsoft better for that.
It changes the playing field completely, and here's how. What is NeXT selling? I would argue that NeXT sells, "We are the computer company that cares about you. We try to build the best products we possible can." There's a level of trust and loyalty that people give NeXT that is unmatched in the industry, and rarely matched outside it. NeXT has that reputation because the company listens to customers. Yes, they make unpopular decisions, and a lot of people hate NeXT. But NeXT customers don't generally feel that way. They generally feel that NeXT is doing the best that it can. Can Microsoft say the same? No.
So NeXT has to make at least a "good faith" effort with this NEXTSTEP port, reflecting the realities of Intel hardware. They have to deliver something they are proud of, that customers feel is worthy of the NeXT brand and the trust relationship that implies. Now, maybe the economics of that work out, and maybe they don't. My point is not the finances, but the brand. The brand is the iconification of the trust relationship. If NeXT can extend that trust to Intel hardware, then Microsoft is in real trouble. Because the only way Microsoft can compete on those terms is by growing a soul.
It is very important not to repeat the same mistakes over and over again, don't you think?