They're the most current available...
X86 went away fifteen years ago. Every "x86" COU built since the last 1990s runs an x86 layer, but underneath is a very different bear.
At the time, the closest the DOS world had to multitasking was TSRs. Beside my first PC was my CoCo 3 with OS/9 level 2 with 512k of RAM with a true preemptive multitasking kernel running on an 8 but 6809 CPU. Microsoft's dominance at the time meant in many ways the most common 16 bit opposing system in the world was only marginally better than a CPM machine from 1980.
I feel it necessary to point out, though, that OS X is not a microkernel system comparable to Minix
While this is true, it's worth noting that a lot of the compartmentalisation and sandboxing ideas that most of the userland programs on OS X employ (either directly or via standard APIs) have roots in microkernel research. OS X is in the somewhat odd situation of having userspace processes that are a lot more like multiserver microkernels than its kernel...
Minix was really the first of its kind; a Unix-like OS that you could run on cheap (relatively speaking at the time) commodity hardware and that you could get the source code for. A lot of the computing we take for granted now comes from Tanenbaum's work.
My first Minix install was on a 386-SX with a whopping 4mb of RAM I borrowed from work back in the early 1990s. I quickly abandoned Minix for Linux once it came out, but for several years I had Minix running on an old 386 laptop just for fun.
I really miss the good old days when technical debates were over the merits and faults of such simple things as different kinds of kernels, and not about whether or not every single thing you do online is being stacked into half a dozen nation's permanent data storage facilities.
The Linus vs. Tanenbaum dustup is from a simpler, more positive age.
That's rubbish. Most of the major platforms have had Java ported to them. Including various obscure systems is ludicrous. If I want a program that I'm almost guaranteed will run without recompile on Linux, Windows, BSD and even many mainframes, then Java remains the best solution. I'm not saying, from a programming perspective, that it's all that great, but from a platform neutral perspective for most of the systems that a programmer will encounter, it remains the best.
Have fun running an x86-64 Linux binary natively on a Windows 8 machine. I can. however, write a Java program that I can almost guarantee will in fact run on x64 Linux or Windows.
No, it is popular because, despite a good many flaws, it remains the best cross platform solution we have.