I think Toadfrog's being a bit of a loveable curmudgeon here! I've also been considered curmudgeonly in my time and, like Toadfrog, I love absolutely traditional renditions of songs without any accompaniment at all.

But...

It's just not true that "traditional singers" were unaware of, or unconcerned with, audience tastes. The fact is that in English circles, Harry Cox and Walter Pardon (no doubt brilliant singers both) are often taken as the norm to whom comparisons should be made. But they lived in communities where a lot of people just didn't care much about traditional songs. Audience taste had comparitively little impact on them because they had comparatively little in the way of audiences! Also, when folksong collectors visited them, they recorded only the most "traditional" (in the collector's mind) songs from their repertoire, and released those recordings that best fit their preconceptions of how traditional singers should sound--the very aesthetic Toadfrog prefers.

Among many traditional singers, though, audience was very important, not because the singer made money but because his status in the community was partly based on his singing. In communities with a high degree of occupational uniformity (i.e. everyone is a farmhand or a fisherman), it is often leisure-time activities that confer a certain degree of status. So singers were often quite concerned with how well they entertained their neighbors. Remember, folksinging is almost always social, a way of sharing art and emotion with others. The reaction of others will always be important, no matter how much integrity an artist may have.

The whole idea that traditional songs should be sung in an "impersonal" or "detatched" way is more a product of folklorists' values than of singers' preferences. Though the impersonal style was an important one in folk communities, it was not the only option open to traditional singers. Once early collectors established this as the ideal, however, singers either conformed to this ideal or were considered "contaminated by commercialism," and their singing was not collected. The ideal was thus a self-fulfilling process.

Personally, I like much of what's done with traditional song and accompaniment. But I reserve the right to dislike anything, too. I respect Toadfrog's likes and dislikes very much, but they are Toadfrog's likes, not based in broad empirical study of folk communities, as he suggests with his anecdotal evidence of Byron Arnold's experience. Byron Arnold, by the way, is not a "recent collector" but did most of his fieldwork in the 1940s. Arnold is expressing an ideal typical of that era, but largely discarded today by academic folklorists.