There is a vast difference in quality between my old Line6 UX2 and my current Audient iD22. But it's not possible to say whether it's all due to the quality of the analog circuitry or how much the actual DACs make a difference.

I definitely heard quite a difference when I switched from my M-Audio Delta 1010 to the Audient iD22.

If you guys keep saying you heard a difference between two interfaces, you might be paid a visit from the Bias Police! 
 
But seriously, as with any type of gear, there's a law of diminishing returns. Once you get to a certain level of quality, which these days doesn't require a lot of cash, improving the sound incrementally gets more and more expensive. Going from a SoundBlaster to an Apollo, for instance, should result in a huge difference. But comparing the Apollo to a Prism or Lavry, any difference might be harder to discern, depending on your monitoring rig.
 
Not having heard the Yamaha console mentioned by the OP, I can't say whether an outboard DAC would be worthwhile. Would be interesting to try, though. See if you can borrow a $2000 Mytek or Bryston DAC, and judge for yourself.
 
Cheers,
Eddie
 

YEA,
I say if you can't hear the diff (between a low end model and a mid to high end model)
then maybe your ears aren't up to snuff!
 

Hi. I heard such big a difference from my old Echo Audiofire 8, to my RME Ucx, that I thought something was wrong.
Of course they were manufactured with almost 15 years between them. Might have something to do with it..
All the best.

I notice a difference between my actual 2 interfaces- focusrite saffire pro 40  ( studio )and the steinberg ur 22 (laptop-mobil).
The difference is not as big as taking the step 10 years ago from my  m-audio 24/96
to the saffire 40.
It´s a good question if  a high end DAC delivers a much better sound as a top 
interface (RME,UAD aso).
Cannot answer that cause I never felt the need for a DAC.
 
Still very satisfied with the sound of the pro 40.
 

Between the converters themselves? Probably not. Just about any converter these days has gone way beyond the limits of audibility. However between the implementation? Possibly. There's a lot of room to do a suboptimal implementation of all the stuff surrounding a DAC. Remember that the DAC chip itself is only one of a number of components needed to get the signal up to line level. Also there can be issues with matching the output to the device you are feeding, and problems getting introduced that way. For example if a DAC has too low a signal, maybe you have to drive the preamps on your monitors harder than they like, and they introduce noise because of it. Or the signal is too hot and you have to attenuate it, but the attenuator has non-linearities at low settings (they often do) and you get a channel imbalance for it.
 
Those are just simple examples, there is plenty else that could potentially happen.
 
So what you are getting with a good expensive, properly designed, DAC like a Benchmark 2 is basically insurance/peace of mind. You know that they did a really good job on the design and implementation, can check the measurements to back that up, and it has the flexibility to be adjusted as needed for whatever it is feeding. You eliminate it as a potential source of issues.
 
In many cases, it won't be audibly different. It certainly isn't something I'd prioritize spending money on (I don't have one), but if you have a really good setup, I can see the appeal to make sure that it isn't an issue.
 
If you do want something like that though, make sure to get one that is verifiably well designed, meaning they have taken, and are willing to show you, the measurements. If a company says it was "designed by ear" and is light on technical details, go elsewhere. We can measure WAY more accurate than we can hear, and we can use measurements to spot problems that only appear in edge cases. I'm a fan of Benchmark for that reason, they provide you with a ton of detail (and you can get more if you ask).
 
Finally you can always try the gold scientific standard of the blind test if you can get a place that'll do a money-back in studio test on the DAC and get a friend to help you. See if you can reliably hear a difference when you don't know which is which. If you can't, then send the DAC back :).

I don't know if it's the DAC, ADC, clock or the op amps or whatever but for the three interfaces I've used most (E-MU 1820M, M-Audio Fast Track Pro and RME Fireface UCX) I can definitely tell a difference. Maybe even identify them with the same recording on all three. The E-MU is pretty decent but rather bright, the M-Audio sounds almost "dead" or "flat" by comparison and the RME is the most neutral with the tightest bass.

So I know these differences exist, even if in this case there is a pretty big difference in both price and manufacturing date. It certainly makes me cautious of using budget devices for anything half serious I'd be doing.

I can say I've heard some differences, but here's one thing to consider.

Take, for example, listening to a great D/A vs. an inexpensive D/A on on a stormy night. While on any normal day, you mightn't hear any real difference, it could be possible that the less expensive D/A might exhibit some unforeseen glitches from lightning.

The differences of good/great D/A converters is usually pretty minimal at normal sampling rates. Clocking is incredibly important for accuracy in transients and for the purest, glitch-free passages. It's even more important for higher sampling rates. Also, the circuitry around the D/A conversion process may color the sound, which is probably more audible. There are, of course, measurable differences which may not always be audible even to the most discerning ears.

There are people who claim they can hear the difference in a $100 vs. $10,000 6-foot IEC Power cord because of the technology involved in the brochure, when they couldn't remotely tell you what quality wiring comes into the building from the street. So, never trust the hype.

As for decision-making, the way I perceive something like that is, consider your client. If it's important to your clients that you have gear they can brag about, and it sounds good, then you did well. If your clients either couldn't tell you how an Apogee stacks up to an Antelope, you're just fine with the Apogee or even a MOTU which I use today. So long as the D/A is clean, accurate, and glitch-free, you should do well.

The quality differences in modern converters between expensive and cheap is minimal if not undetectable.
Unlike the 80's where a lot of even expensive converters were crap.