If you want something easy to simply turn it on and start composing music without having to fight your instruments, I think note-performer is a good investment. It’s easy to use, and won’t shred your ears when you’re stuck in a room with it all day.
As for more musical and realistic mock-ups…we tend to crank these up while mixing, and the annoying flaws of ANY library can be exposed (they all have them). It’s the stage where we compromise, substitute, and spend as much time trying to remove/cover/substitute than we do thinking about how a ‘real musician with the real acoustical instrument would be doing this phrase’ (often using tricks that won’t be on the score, or are even quite different from what the written page says ‘in theory’ a ‘human musician on stage would be doing’).
For me, it’s still easier to not worry too much about the sound in Dorico (Or Sibelius, or Finale), and move to a DAW when the time comes. It would take me pages to list all the reasons why; so, the short version is this. When I want to compose and collaborate about the score itself, what matters is a setup that doesn’t disturb the creative flow, and goes easy on my ears. When composing, I’d rather have something that is clear at reasonably low volume, yet pleasing enough to sit there all day and work with it than have it be ‘realistic sounding’.
That’s often NOT what you want when doing a mix-down that you intend to master for distribution, that could get played in everything from cheap ear buds to tin button cell phone speakers. The libraries that are often THE BEST for getting a realistic mock-up, can shred your ears to bits, or sound rather ‘harsh/nasty’, during long work/composition sessions.
A top of the line DAW with top of the line features for working with virtual instruments was designed specifically to engineer sounds, with tens of thousands of bells and whistles especially for controlling every minute detail of what comes out of the speakers as quickly and efficiently as possible.
However, It’s not so great at notation (or attempting to interpret what we want it to sound like on its own). It’s not always the most inspiring thing to compose in, and have an ongoing working model to collaborate with others at the drop of a hat though.
Having said that, Dorico is getting much closer with each release to being a total one stop shop for working with virtual instruments. The playback feature sets have been taking rather significant strides with each release. Instrument libraries are also getting better, and their designers are becoming involved in helping them hook up with with the VST3 age and beyond.
Dorico is already quite good for communicating, collaborating, and getting your point across. If you’re willing to put a lot of time into tweaking the score, and the wheels and dials of your virtual instruments for optimal playback, it can already be done fairly well in Dorico, but in my opinion, not nearly as quickly and efficiently as in Cubase (assuming you know Cubase/MIDI/VST/etc. pretty well, and are making the best of its instrument and audio shaping features…so many details that can be done with controllers in real time, or with logical editors and such in minutes or seconds, that often take HOURS to sort out by hand with your mouse and bars/dots/lines on a screen).
In either case, there’s going to be a learning curve to getting realistic mock-ups (no matter how much you spend on instrument libraries). MORE than half the battle is in mixing, and dynamic control.
If working with traditional scores is essential to your work, or even if you simply find that it helps your creative flow to work that way, I don’t think you’ll be disappointed in Dorico in the long haul. Get your hands on it as soon as you can and start learning it.
Beware of the pitfalls of trying to be both a composer, and an audio engineer. Both are very time consuming, and require a different set of listening skills and flavors of patience/tolerance. The mix one goes for when he is sitting there composing all day is often WAY DIFFERENT from what you’d shoot for on something you plan to distribute and might be played on everything from ear buds to tin button cell phone speakers. It’s far too easy to ‘learn’ to like things that sounded good to YOU while sitting in that chair, in front of YOUR speakers, yet it sounds ‘terrible’ to people who haven’t been parked in front of your monitors with you all that time, ‘learning to like it’. Kinda like working a high school orchestra with terrible instrumentation, in a bad rehearsal room. Your brain plays tricks with you, that aren’t going on in the heads of others, especially the second you step into another room.
So…maybe it’s just me, but…
Keep it simple when composing. If it’s warm and fuzzy enough that it’s not shredding your ears, yet clear enough that you can pronounce the important phrases and ideas with relative ease, then you’re golden.
When you get ready to do a mock-up, change gears. Think musically. CHEAT! You can use tricks that you’d never put on the score to trick brains into thinking it’s more real sounding than it really is (layering/doubling/room noise track/etc.) 
Digital instruments and rooms don’t really work like real instruments in real concert halls and studios, and every set of speakers is different. In time you’ll learn that compromises or psycho-acoustical trade offs often have to be made between ‘realism’ and ‘musical/pleasing’. After all, you’re trying to get an orchestra sound out of paper disks…