"The Rubsam approach seems to me to make the music sound ancient, a keyboard analogue of a Gombert or de Rore motet, but more rhythmically free. Each variation is a motet for keyboard ... "
I've only listened to sound clips so far, but had the same thought. The tranquility of Rubsam's playing and mellow sound of the lute harpsichord reminded me of a Renaissance motet that has been arranged for lute duet (as was popular during the middle to late Renaissance). I even started to forget that he's playing a keyboard. Which is intriguing, since I've long thought that Bach's music had strong roots in the Renaissance. Not only do I suspect that Bach maintained an intense interest in Renaissance polyphony throughout his life (as you may recall, I think there's a possible connection between Bach's Art of the Fugue and Josquin's innovative late Missa Sine nomine: ), but also in Renaissance practices as well, such as the common use of one singer (or two) to a part--a tradition that Bach's idol Buxtehude later adopted, along with his friend Telemann (as did Bach himself).
What gives me further pause here is that the lute harpsichord is a very rare instrument. I understand that there are only a few in existence today (Rubsam's being a new instrument built by Keith Hill). Therefore, lute harpsichords couldn't have been widely used during the Bach's era, and yet Bach owned not one, but two lute harpsichords at his death.
After listening to the instrument, I can't imagine that Bach wouldn't have preferred it to a clavichord, as the lute harpsichord doesn't have any of the drawbacks or limitations of the clavichord (i.e., difficult to control and play in tune, etc.). Yet it offers the same intimate, mellow sound. It also has a fuller, more resonant sound than a clavichord and appears to be capable of a greater subtleties. So, how could Bach not have preferred it?
As to whether Bach considered it a suitable instrument to play his Goldberg Variations on, if we trust Johann Nikolaus Forkel's 1802 story that the variations were composed to be played on a harpsichord by the teenaged Count Goldberg for the late night listening of the former ambassador to the court of Saxony, Count Kaiserling (who was ill and had insomnia), the idea that Bach privately played the variations on a lute harpsichord seems unlikely. Although the mellowness of the instrument is arguably more suitable for the intimacy of late night listening than a harpsichord. It would be interesting to know whether the Counts Goldberg or Kaiserling actually owned a lute harpsichord. Of course, Bach does specify on the title page of the Goldbergs that they are intended for a two-manual harpsichord. Therefore, we know that the Goldbergs weren't composed for a single harpsichord, or clavichord, or lute harpsichord (unless there was such thing as a two manual lute harpsichord?). In addition, I have doubts about whether Rubsam's approach would have worked as effectively if he'd taken all the repeats, & Bach does ask the performer to take all the repeats.
What's refreshing about Rubsam's approach is that it's the total opposite of the dogmatic, tightly controlled 'academic' approach to Bach's counterpoint that you find with a lot of the older harpsichordists. Indeed, the rhythms are so free that he seems to almost displace musical lines--in a fugue-like manner--then bring them back together again. The effect is dream-like. However, I don't think this approach would work as well on a piano, as it seems more suited to the inherent qualities of the lute harpsichord. Rubsam's ornamentation also works quite well on the instrument. I felt drawn in. It added interest, & made me listen more keenly.
I was also struck by how, even though Rubsam tends to pull the music apart, his playing isn't at all romantic. Which brings to mind pianist Elizabeth Rich's rhythmically free and varied approach to Mozart's piano sonatas. I look at her approach as very classical, though others may not. Like Rubsam, Rich puts the content of the music ahead of other concerns, and her understanding of the style of the period is so strong that her playing doesn't come off as 'romantic' either, even though she brings a greater rhythmic freedom and contrast to this music than other pianists. Considering the illuminating results, I suspect that Rich's approach is closer to the way Mozart played his own music than any other pianist. As with Rubsam, she draws the listener more intensely into the beauty and meaning of each phrase. There is a sense of time slowing down, which allows the listener to linger and reflect more deeply on the character and meaning of each phrase.
"They all had pupils who followed their ideas. There's no Rubsam "school"."
That's surprising, since Rubsam is a teacher. I recall that he teaches or has taught at Northwestern University. On the other hand, it's not surprising, since in order to most effectively adopt Rubsam's approach, his students would have to play a lute harpsichord, in my view. Personally, I find it hard to separate what Rubsam is doing interpretively from the instrument he's using. With so few lute harpsichords in existence, in order for there to be a genuine "Rubsam school", Keith Hill would have to get busy with many new commissions, & that's not likely going to happen. Although it would be interesting to hear some alternative interpretations of Bach's keyboard music on a lute harpsichord, considering that Bach owned and played the instrument.
I see the Brazilian organist Julia Brown studied with Rubsam at Northwestern. She has recorded the complete organ music of Buxtehude for Naxos (& now lives in Oregon). Have you heard her playing? If so, do you find that her Buxtehude has anything in common with her teacher's Bach?