In 'Managing Technical People', 1997, page 199, Watts Humphrey says that, after several failed attempts to produce a PC by IBM procedures, they set up an independent team that could skip the procedures as necessary to get the job done. This worked in the short term but it had two side-effects that were catastrophic in the long term: they lost control of the operating system to Microsoft, and they also lost control of the chips to Intel. He says both of these side-effects would have been caught by the checks inherent in the normal IBM procedures.

They might have caught the mistakes but I think the resulting product would have been a flop.

If IBM hadn't done what it did somebody else would have dominated the market with a product to fit the same niche. Perhaps somebody "downmarket" in the more consumer space who managed to punch upwards -- maybe Apple who had some business success with the Apple II + VisiCalc, etc. or maybe Kaypro or somebody in the CP/M space. Or perhaps somebody else "upmarket" like DEC, who came too late the personal computer space with products that nobody really bought (DEC Rainbow, etc) but maybe they'd have had more success if IBM hadn't gotten in there.

The market wanted a relatively open product to innovate in. When the PS/2 came along a few years later with proprietary bus, etc and tried to put the genie back in the bottle, it flopped.

One of my favorite scenes from Pirates of Silicon Valley is when Steve Ballmer (played by John DiMaggio, yes, the voice of Bender) narrates an audience aside, positively giddy with Ballmer glee at how his clever friend Bill put one over on those stodgy fellows from IBM and got them to give away the golden goose. It was like harpooning the great white whale. While it was a telefilm and thus a fictionalization, near as I can tell things happened pretty much as described.