"Efficiency is intelligent laziness." -David Dunham
"Think for yourself!" - Abigail
"Memory, processor, disk in that order on the hardware side. Algorithm, algorithm, algorithm on the code side." - tachyon
In my experience, Linux (for example) is at least a lot closer to meeting every one of those goals than Windows. Obviously, these criteria are somewhat subjective in real-world application, but if you use either Linux or Windows as a basis for comparison, the other falls short of (in the case of Windows compared to Linux) or admirably meets (in the case of Linux compared to Windows) those goals.
Granted, I don't know enough about the code in the Linux kernel to be able to comment particularly on its reusability and readability, but the fact that it's open source makes it reusable and readable, even if only with difficulty, while the closed source status of the Windows kernel makes it unreusable and unreadable — to say nothing of the fact that it's such a bloated, crufty collection of layered kludges that I have a hard time imagining pieces of its code being terribly reusable or readable, anyway.
Don't redefine the terms. The OP was talking about reusable code in terms of software engineering - not about about licensing issues. perl is open source, but it's not very readable or reusable. And I've seen a lot of code that's both readable and reusable according to the definitions the OP gave, but which isn't open source.
1. I can't read his mind. I only know what BrowserUk said, not what he thought. As for the original poster, it's true that he was referring to software engineering as opposed to licensing, but see #2 below.
2. I made reference to licensing issues, as I tried to point out, only in absence of direct personal knowledge of the software engineering characteristics of reusability and readability.
3. I tried to indicate clearly that I referred to Windows and 'nix OSes in comparison with one another, and not as stand-alone Platonic ideals. One cannot really treat an OS as an absolute when discussing its design characteristics (not if one wishes to be taken seriously, at any rate), and OSes must therefore be examined in the context of one product among many. That's all I did.
I know you'd rather I simply say "Well, since I don't know everything, I must be wrong and you must be right," but I'm afraid I have a tendency to actually try to approach discussions with a desire to see valid arguments made rather than rely on the emotional impact of logical fallacies to make my decisions for me.