The cache-busting works by appending an always-unique query string to the end of every file which is required.  It makes use of RequireJS's urlArgs config value; RequireJS takes care of appending it for you:
urlArgs: "bust=" + (new Date()).getTime()
The(new Date()).getTime() part is just a simple way to get a unique string out of JavaScript.  You could do some variation on Math.random(), but using the number of milliseconds since the epoch guarantees uniqueness, for optimum cache-bustage.
I think Mr Burke is suggesting something like:
require.config({
    baseUrl: '/base/path',
    paths: {
        'fileAlias': 'fileLikelyToChange?bust=' + (new Date()).getTime(),
        'anotherFileAlias': 'anotherFileLikelyToChange?bust=' + (new Date()).getTime(),
        'jQuery': 'jQuery'
    },
});
So, instead of the ubiquitous urlArgs cache-busting, you apply it specifically to each file which is likely to change; hence, excluding any libraries.
I haven't tested it, but I'd probably tidy it up to something like:
function bust(path) {
    return path + '?bust=' + (new Date()).getTime();
}
require.config({
    baseUrl: '/base/path',
    paths: {
        'fileAlias': bust('fileLikelyToChange'),
        'anotherFileAlias': bust('anotherFileLikelyToChange'),
        'jQuery': 'jQuery'
    },
});