Rudi has changed his opinion a bit on the nature of feedback in this amp.
have built scientific instruments for many years, teached and written books, so I know what is the mean of the kevin's design opamp ( I used simple words with you since I dont' know your skills ). I use "DC servo" on many instruments but not in audio, since in real word a zero is never a zero and a silence is never a silence ( I use your terms ) the opamp always act as a feed back generator ( as secondary effect I admit ).the ammount of feedback should be small but not negligible from my point of view . I believe to talk about hi-end and not "basic audio", and in my hi-end way nothing is to small to be negligible
From my poimt of view the the advantages of this totpology ( no output caps ) are less then disadvantages, and in a deep analysis it must be considered as feed-back circuit.
About input stage. The original article is interesting and well written, and I agree with author when he say the circiut works in a similar way to long-tails. Well I consider long tail as a internally feed back topology ( but on this I admit is possible to have different poit of views, since the feed.back is not so evident).
Obviously kevin and me are designing in different way, he is a skilled diyer using the component manufacturer databooks standard circuits and well known topologies, very good for industrial production, less for hi-end research. (IMHO)
I use my experience, my lab and my finance to go deeply and found smething different.
rudi