Re: IBM have gone the way of Oracle..
That was always their plan, you are just a 'user' of compute resource, you don't need to worry about what it actually is, or how it actually works. You just need to be able to understand how much compute, I/O, and storage you require. IBM will take care of the rest. Welcome to the world of the virtual computer and the virtual datacentre, where you only rent capacity from companies like IBM.
Try to imagine a world in which you manufacture lots of different systems with incompatible architectures. Each providing a range of different capabilities in terms of compute, I/O, and storage. You sell your systems to lots of different companies (one time fee). Some of them sign up to have you provide services for a fixed term (one time fee, only repeatable if they don't go and sign with a different company).
Wouldn't it be much better if you only actually rented capacity? Where you had a regularly repeating income stream (because moving to another provider was a damned sight more difficult for people who don't actually have any hardware, or capacity to store their own data). Those companies would save lots of money buying hardware, specialist skills, staff, etc, and they'd pay it all to you. It'd be much better because once they're with you, you've have an open door to sell them whatever you want, services, and solutions are you bread and butter and all your clients are going to need lots of them now they can't do anything for themselves.
Of course you should plan such a take over the world strategy very carefully, what you don't want is to spend vast amounts of the revenue stream on providing the capability/capacity you have to provide. Now if you only had one operating system which could run on all your disperate architectures, was powerful and capable enough (or could be developed to be so) into running as a virtualised parallel system across multiple physical systems of different hardware architectures.
With such an operating system it would mean you could have a mainframe, and a couple of smaller X86 type systems provding a system image with enough capacity/capability to suit a specific requirement. Upgrading would be easy you could add a system with about the right amount of compute, I/O and/or storage to the system. Such a system would cost a fortune to develop of course, unless of course you can get some of that done as 'good will' in return for your own 'good will' in improving an operating system others could use, not to mention the additional capabilties you would add to it because you need them for your take over the world plan. This is why IBM leapt into Linux, Java... etc.
Before anyone kicks in with comments about my use of the words 'take over the world', all companies have a 'take over the world strategy, be they Google/Microsoft/Oracle/IBM. Only they're very canny at IBM, they've been running with this strategy for more than 15 years that I know of, and hey presto here we are living in the age of 'Cloud'.