This is still being argued at the present time (2001). Wilhelm Rust (BG) still considered the cantata mvts. with obligato organ to be the source for the harpsichord concerto, but now there is no doubt that the cantata mvts. (BWV 169/1+5 and BWV 49/1) were based upon a now-long-lost solo concerto with a concertante instrument (ein verschollenes Solokonzert mit konzertierendem Melodieinstrument) (Ulrich Siegele). There is up to now no general agreement on what this instrument might have been and what the original key of the composition was.
Ulrich Siegele (Kompositionsweise und Bearbeitungstechnik in der Instrumentalmusik Johann Sebastian Bachs) (Dissertation, Tübingen, 1957), published 1975 Neuhausen-Stuttgart (Tübinger Beiträge zur Musikwissenschaft, Bd. 3) p. 137, thought that he could, with some degree of probability, see this as an oboe concerto in Eb Major. As reasons he gave the relatively narrow range of the solo part (a little less than 2 octaves) which would be unusual for a string instrument and the “Umbruch” [Oktavknickung?] of the 2nd ripieno violin part in m 9 of BWV 169/1. These considerations led him to consider the possibilities of a Transverse Flute concerto in F major and and Oboe Concerto in Eb Major, where he considered the latter to be more probably than the former because ”the connection of the Oboe with Eb Major is smoother/easier/more common (geläufiger) than that of F major for a transverse flute.
Wilfried Fischer, in his dissertation (Tübingen, 1966), agreed with Siegele and provided a reconstruction of the Oboe Concerto in Eb Major. He found further ‘proof’ a series of errors (all of them off by a 2nd from the intended note) in mvts. 1 and 2 as well as in the cantata mvts. in D Major and B minor (BWV 169/1+5).
In his NBA VII/7 volume of reconstructions, Fischer decided not to include this reconstruction since he reinvestigated the situation and came to the conclusion that “in regard to the key of the composition and the solo instrument used in the original composition, many questions still remain, although a somewhat reliable reconstruction might be possible.”
Persuaded by Siegele’s solution, Joshua Rifkin, in 1982, published his own reconstruction as a recording (Pro Arte Digital 153 (1983): Soloist: Stephen Hammer). In contrast to this, the NBA editor of BWV 169 in the NBA KB I/24, Matthis Wendt supported the following argument: “It is a fact that corrections of errors off by a 2nd are relatively common, but the number of errors committed in BWV 169 are insufficient to come to the conclusion that the original concerto source must have been in E Major or Eb Major. The NBA editor for BWV 49, Ulrich Bartels, agreed with this assessment.
In 1983 Arnold Mehl published a reconstruction in D Major for Oboe d’amore as the solo instrument. Bruce Haynes, in 1992, supported this reconstruction without making any reference to Mehl’s work.
Recently, Wilfried Fischer attacked this problem once again [the search for the type of solo instrument used] and considered that the choice of key was without any doubt Eb Major, but that the choice of instrument should be a viola. For this, one would have to assume “that Bach had transposed the part from the original instrument up one octave” – this being, however, a procedure for which there is no precedent in Bach’s oeuvre.
Based upon its use in the sacred cantata, the original source (solo concerto) must have been in existence before 1726 at the latest. Gregory Butler (“J. S. Bach’s Reception of Tomaso Albinoni’s Mature Concertos” in Bach Studies 2, ed. by Daniel R. Melamed, Cambridge and New York, 1995, pp. 20-46) thinks that it is possible that this original concerto was composed under the influence of Tomaso Albinoni’s Concerti a cinque, 1722 and suggests that the Bach’s concerto was composed during Bach’s early years in Leipzig.