I think digital is overall slightly cleaner than analogue where the room and the circuits have automatic slight saturation or something to them.
I think digital synths are the true sound while hardware is just more saturated alittle.
I also have a preconceived notion that hardware has overall more natural saturation since there's physical circuits and whatnot.
Watch this video:
https://www.youtube.com/watch?v=xRIxvpdZ5Rs
This is just a glimpse into the sound of digital. It's not software that is cleaner, it's hardware that is cleaner.
Remember that there are DAWs like Cubase out there, where you cannot even see the indications of this kind of stuff.
So you basically need the ears of a dog in order to combat all of the digital noise that is morphing into the signal under the hood in various configurations.
But besides this, there is a lot of noise also introduced by the digital algorithms themselves, both directly and indirectly. This is why you need to be incredibly selective about what plugins and what plugin configurations to avoid. In practice you end up using plugins that eat on the signal. And then when you use things like routing and 8 times oversampling to minimize this, then you get added phase instead by the increase in system load. So, the better you want to make it sound, the worse sounding it might get. And I know systems that degrade the sound over time, because of memory leaks in the DAW, the drivers etc. So it's not only that you have various things you can deal with, some of the noise finds its way into the signal due to various types of system related issues in your particular mix of software and hardware. In short, pros must limit this kind of stuff from happening, and the way that is done is by offloading the system as much as possible by using better sounding hardware instead. That improves the resulting sound cumulatively, because it positively impacts on the signal both directly and indirectly.
And related to digital software based music creation, there is some funky stuff going on. For instance it does not make any sense at all that music that does not clip at all when you play it inside of the DAW, might clip by almost 2 dB in a more lossy format. Sometimes I wonder if there is some evil force out there that does all of this on purpose. Now, I compensate for this with my masters, but I know many engineers that don't. With hardware in the signal path you might be able to afford this because you have things like harmonic distortion that the perception is influenced by as well. But with software you might not have that to the same degree. So it's like no matter what angle you approach it, hardware ends up winning.