Really the way to look at the Linux paradigm from a purely "make it work" viewpoint is this...
Think of a photocopier or a self contained Casio keyboard or whatever. It has it's own little code written onto chips (or even little hardrives) and all they do is make the device work. It is programmed to do exactly what it needs to do... no more, no less. It does not get interrupted by other conflicting instructions. None of its resources get hogged by other tasks. It just works.
With dedicated distros those instructions are on your terminal (your computer) and send the data where it needs to go to make things happen. If you don't add any extra gack to the process then the terminal just focuses on those tasks and all the hardware involved does not get distracted.
With Windows or OSx or even Linux distros like Ubuntu that are intended to "do it all" and mimic those other OS's it has a million and one things going on. It is always ready and waiting to handle whatever you throw at it. Which is great for multitasking.
But it also takes a lot more resources, causes a lot more potential conflicts and interruptions and is just generally unnecessary if all you want to do is ONE specific task.
We don't install a full consumer version of Windows on a photocopier because that would be silly.
Servers are kind of the same way. You need more than a simple photocopier needs but far less than a system that someone is going to do "whatever" on... but you also need more in the way of data management/storage and intrusion protection. You want your terminal's resources to go to those tasks instead of all the crap a modern consumer PC does. You aren't going to be watching videos on it. You aren't going to doing a bunch of spreadsheets or editing or whatever. You just want it to retrieve and serve up data and thwart incoming threats.
Similarly... with audio what are we doing? It's far more complex than a web server and certainly far more complex than a photcopier BUT we see it day in and day out here on the forum. Windows settings meant for generalized usage are interrupting data flow. Background programs are interrupting data flow. Bad programs/drivers conflicting with the OS are interrupting data flow. Yadda yadda.
I showed up here over three years ago and gatdamned if it isn't a massive annoyance trying to keep track of all this crap you have to do to make sure all this OTHER stuff that is really just designed to give a more generalized audience a GENERAL release that will serve MOST people's needs isn't interfering with our very specific needs as music producers.
In a perfect world we could get Microsoft to release a version that caters to us and is completely optimized for audio but no way in bleeping poop is that going to happen. They won't let us crack it legally... so what do we do?
Well, we rely on guys like Mr. Roseberry to build and tweak our systems and that is brilliant BUT having streamlined OSs that simply remove all that extra crap would be great. Then Jim could focus on testing hardware... lol.
This was I think part of the whole reason PT was the market leader for so long. Their dedicated boxes handle all that nuisance and took much of the OS out of the equation. Same for video. You just couldn't rely on the core system and OS to deal with that crap and you paid through the nose for the luxury of being able to avoid it. Now our systems are able to deal with it but our OSs are so convoluted and scattered we have to crap our pants trying to worry about all the fiddle faddle.
I mean it's not as bad as all that and if a moron like me can make things work we truly are in a golden age but really if the code was there to just make sure all the data was going where it needed to, the common platforms (like VST) were available, driver support for the main manufacturers were in place we could be running MUCH more on MUCH less.
Dare to dream... eh?
Sorry for the pontifimication and excessive elaboramations... weird mood today I guess but damned would it be cool to just put all that crap behind us.
Audio terminals, baby!!!