Yes. What measurement indicates pretty much every sigma delta DAC to be grainier and rougher in treble than R2R DACs? Let's look there or find it first rather than beating around the bush.

On the topic of sigma delta stability, loop filter and noise shaping and etc, it's a very complex subject:

http://yadda.icm.edu.pl/yadda/element/bwmeta1.element.baztech-article-BUS8-0020-0008/c/httpaa_czasopisma_pan_plimagesdataaawydaniano1201108noisetransferfunctiondesignandoptimization.pdf

How are you going to determine which noise shaping works best? It's kind of a black art, like how ESS tuned their Sabre dacs.

http://www.teddigital.com/ES9008B_tech.htm

Disclaimer: I own both the Audio-gd master 7 and Schiit Yggdrasil, both are R2R DACs.

I haven't heard any PCM4222 or AKM1234 before. So I can't make any comments on them.

The hybrid sigma-delta/R2R Ti PCM179x is only mid-fi grade at best. Can be dull/laid back sounding. It's predecessor, Burr Brown/Ti PCM1704UK is so much better in all aspect.

Certainly there are are good sigma delta dacs out there. AD1955(Bricasti M1) or AKM4399(Gungnir) or Matrix X-sabre(ES9018S) or Geekout V2(ES9018AQ2M/K2M) or the NAD M51(propriety sigma delta).

However these dacs are certainly not nearing Yggdrasil or MSB Analog kind of performance.

Go listen to one of the good R2R dacs for yourself. Measurements and Specifications are only part of the story, the key performance benchmark of an audiophile grade dac is level of sensory enjoyment you get out of it(subjective).

Certainly there are are good sigma delta dacs out there. AD1955(Bricasti M1) or AKM4399(Gungnir) or Matrix X-sabre(ES9018S) or Geekout V2(ES9018AQ2M/K2M) or the NAD M51(propriety sigma delta).

However these dacs are certainly not nearing Yggdrasil or MSB Analog kind of performance.

But Yggdrasil is a PREMIUM dac with lotsa stuff (improvements) OTHER THAN THE DAC CHIP.
That was the whole point of the Counterpoint DA-10 ref. I made earlier in the thread.
You've gotta level the playing field. (The ancient Counterpoint D/A does that to a very large, scientific extent, using its proprietary module)
That said, a trained listener can distinguish acoustic differences among electronic architectures and schemas: e.g., non-oversampling vs. oversampling. But I think the value-judgement calls between D-S vs multi are much trickier. I've gotta a closet full of vintage multibit (including some high end stuff) ... yeah it sounds good/different ... but not important enough to drag outta retirement. My daily work-and-play D-S system -- Asus Xonar ST-based -- is aged and clunky, but no way I'll trade in the convenience of multi-TB HDs and playlists with months full-o-FLACs.

Go listen to one of the good R2R dacs for yourself. Measurements and Specifications are only part of the story, the key performance benchmark of an audiophile grade dac is level of sensory enjoyment you get out of it(subjective).

"Measurements and Specifications [as known to human science mid-2015] are only part of the story". Science and metrics will improve, and parameters will be added to the growing list of tracked specs. In the end, it may not matter ... if you sample enough in the A/D process, you don't need fancy D/A. That's essentially the principle behind DSD; Google "The Best DAC is no DAC".