In
this article by Briess (written by Dan Bies, who I don't think is a lady), they looked at color extraction and "total solids" extraction for various malts in a 60 minute Cold Steep vs a Congress (mash) wort. Since the process being discussed in this thread is mainly about crystal/caramel and dark roasted malts, I'll just include their data here. Each cold steep percentage yield is given as a function of a Congress wort yield, i.e. think of the Congress yield as 100%.
Carapils: 95% of color, 32% of total solids
C20: 95% of color, 49% of total solids
C120: 77% of color, 43% of total solids
Chocolate: 69% of color, 70% of total solids
Proponents of cold steeping specialty malts, especially dark roasted malts, often talk about reduced astringency/bitterness as the reason for doing it. Some who have tried it say that there's less astringency/bitterness, but also less color and flavor. I think this is all consistent with the Briess results. Note the numbers above for Chocolate malt (the only dark roasted malt tested). With the cold steep, there's considerably less "stuff" being extracted from the malt. But... since you're getting less color and flavor, how do you fix that? Well, you can increase the amount of malt used, thus also increasing astringency/bitterness (as compared to the lesser amount of malt). Is there a point where the color and flavor are equivalent to the mashed wort, but astringency/bitterness has still been reduced overall? Maybe. Cold steep fans would probably say yes, because the temperature was lower and the pH would have been lower (compared to a normal mash), so less tannins were extracted. And they may be right, but
more malt was used, so who knows?
The above discussion applies to cold steeping, but I think it largely could apply to late addition of malts to the mash and to "capping" the mash just before sparging. i.e. processes that extract less stuff, which can be compensated for by using more malt. Is there really a net benefit, and if so, is the juice worth the squeeze?