Should I upgrade from a Mac Mini to a dedicated server or should I upgrade my dac?

@pgastone

@jjss49: I am sure that everything ultimately matters to some degree but I like to move incrementally and see how the system evolves. And since there seem to be various approaches I am just trying to get a sense from others who have more experience and expertise than I do on where they think is the best place to start.

you can start wherever you want, but based on the component list you shared in your initial post, and assuming it is set up right, your system should definitely be good enough to hear the sonic advantages of each improvement step (streamer, dac, core/network)

in my own experience, i did the dac upgrade first, it was kind of an obvious place to start... but the streamer upgrade and core/network purification/filtering steps were each quite significant in sound quality... if put gun to head to put numbers on apportion degrees of improvement to each step, i would say 50% dac, streamer 20-25%, core/network 25-30% -- so given that, if you want the full monty, you can’t really skip any of the above, especially as the improvements are cumulative in nature

longer version of what i was trying to say earlier...

You could benefit from dedicated server vs stock mini, but with no rendering things aren't as bad as they could be. Stock mini is noisy environment for a streaming component so there are some losses you'll never recover further down the chain, this even though doing no rendering. Certainly, there are nice benefits with a quieter network, but optimizing rendering is bigger bang for buck. You could continue with a streamer/dac combo or separate the two, either way has potential to be greatest improvement over what you have now.

Personally, I think you're on right track not doing any rendering within server, continue this into future. Still, after upgrade of dac or streamer this may become weak link in chain. Your mini is noisy, could be bested with dedicated server, but cost/benefit ratio is relatively low since you're not doing rendering. If I was in your position at this point, I'd think about bypassing switch with server with two network ports, second network port goes directly to streamer or streamer dac. Benefits would be twofold, simpler more direct network path best for eliminating noise, dedicated server with lps quieter environment vs mini. As I mentioned above, any dedicated server WITHOUT TWO ethernet ports would provide much less benefit.

The Roon Nucleus wouldn't provide enough benefit being designed as rendering server, and would need lps for best performance. Of the SMG, only the SMG I9 optical would be substantial upgrade, and here paying for capabilities you don't appear to need. Custom built or Antipodes K series server way to go down the road, server designed to be partnered with streamer is what you're looking for.

Moving the ROON core from my Mac Mini to an SGC ST was the right move for me. Already had a Pi based streamer and a descent DAC.

I can't thank all of you enough for this great advice. 
This is great!
Based on everything you mentioned right now I am leaning towards my initial impulse of starting with a better dac and implementing the other upgrades over time,

I like the idea of integration but only to a point- as appealing as a single box solution may be I'd rather retain the ability to upgrade single components .  That said, I would probably opt for a dac with a bridge/network Roon endpoint- I was thinking of a Bricasti M3 or maybe an M1 SE (really stretching budget here); another audiophile I know well swears by the Manhattan II.  I believe any of these these should be a good step-up from the Moon 280D (though it remains a very competent dac).
The other option of course is a dac without any network card (maybe Aqua La Scala or Holo May) and investing in a separate Roon endpoint- Sonore etc as discussed above.  In other words, spending less on the dac itself and to give more room to an endpoint upgrade. At this stage I'd probably leave the Mac Mini for last in the upgrade process.

In all of this I should mention what I am looking for; first and foremost I would like to eliminate that electronic hash that bothers me, and second I'd focus on a believable soundstage.  Many often refer to soundstage width as an essential element but given the choice I tend to care more about depth,.  And as I have experienced with analog when properly set-up, that sense of space goes hand in hand with "forgetting" about the speakers and all the equipment upstream.  For this I can also forfeit some microdynamics for a better sense of the overall musical event (I sound like an analog person! but don't want to go overboard either as I do care about transparency, detail...and dynamics).

@pgastone that makes sense! Good luck and enjoy the journey!