Frankly, I think this kind of thinking is short-sighted quarterly corporate bullshit almost certainly done by IBM and Redhat itself would never have done it (and most of the people who were Redhat in those days left to work somewhere else). How many corporate environments the world over are running on Redhat, paid Redhat, because the engineers there came up in the Linux world learning free Redhat and were able to continue using CentOS at home or in their dev environments? That sort of cradle-to-grave brand identity with people who eventually become high-level decision-makers and money controllers is invaluable, but the value is also impossible to quantify, accountants never se it, and instead just see people trying to use all of their hard work for free.
Yet somehow even Pablo Escobar understood this and threw away billions just giving money to poor people to keep cities on his side. I'm not even going to say IBM's basic business model of making shit that works, doing at least some complex things reliably and well back when that was very difficult, is invalid. But it is fundamentally different to Redhat's business model, which was simply cultivating a lifelong love of their product in the minds of its users. It's like the Army acquiring the Red Cross. They work in a lot of the same places doing overlapping work and no doubt a lot of corporate synergy could be achieved and administrative overhead eliminated, but it would irreparably destroy a century and a half of goodwill and brand identity. Exactly the kind of strategy you'd expect from people whose decision consequence horizon extends no further than the next three reporting periods.