Can anyone confirm that there are so many trills in the score of the
Sarabande from the lute suite in e (BWV996) as Robert Hill does on his
lautenklavier? I didn't notice such HUGE abundance of trills in any other
performance of Bach.  I can even say that the melody itself is hardly
distinguishable under the thick cover of trills.  I listened to Segovia's
version on the guitar to compare it and, although Segovia barely connects
sounds and sounds too Spanish, the melody is much better heard.  The
melody's so good that I'd rather remove most of the trills than stick
to authenticity.

As a side note, what do you think of the guitar recordings of the
lute suites in general? For me, the guitar sound is very fragmented,
chopped, having little resonance and the music cannot come out in a
smooth rivulet-like flow when it would be nice if it could.  The guitar
is a wonderful instrument in the agitated Spanish music, including Gipsy
Kings, but I'll probably forget about guitar versions of the Bach's suites
and limit myself with the lautenklavier.  The instrument sounds so good
(at least in the Hill's nicely echoing recordings) that it seems more
appropriate in other works as well - even more appropriate than the
harpsichord that the works are usually played on (take Hill's BWV906
fantasia, for instance - I now can't imagine a better sound for this
jolly piece than that coming from the lautenklavier).

Juozas Rimas Jr (not the one playing)
http://artists.mp3s.com/artists/141/juozas_rimas.html