I was just idly wandering the web last night when I came across the Classix II and decided to simulate the Dayton 160's in VituixCAD enclosure designer and Paul Carmody has indeed chosen the optimum cabinet size of 15L for those drivers but a 15L bookshelf is a big bookshelf but a small floorstander so I won't be making any of these soon. (You would have to put them on stands and if you do that you may as well make floorstanders)
However it's caused a paradigm shift in my thinking and I now realise that one firstly decides a rough idea of your cabinet size based on your application (computer, bookshelf, stand mounted, floorstanding or wall built in) and then choose the driver. Each manufacturer will carry in its range a driver which is best suited to a given cabinet size and all within the range will be of roughly the same quality. To read in some forum that woofer xyz is great so you should use it is to put the cart before the horse. I'm saying this with the caveat that you know how to set the crossover frequency before higher cone modalities kick in.
Getting a decent tweeter is a different matter altogether, especially getting one that has a resonance low enough to set a crossover frequency that suppresses modalities. There are an enormous number of tweeters that simply don't cut the mustard either in having a low enough resonance or a decent high frequency response. Look carefully at their published frequency responce and pay attention to NEGATIVE comments about any tweeter you're considering. Be prepared to pay much more for the tweeter than the woofer.
Playing with VituixCAD has also made me realise what a nonesense any given wisdom is about choosing closed verses reflex based on Q(tot) is. In this day and age with tools like VituixCAD and WinISD available there's no excuse for getting the enclosure volume and port dimensions wrong, or at least not understanding what the compromises you've made are.
P.
