Dear Carey.
I am afraid I would want to be a little careful with dismissing Paul’s and Elaine Fuller’s assertion. I think I am well familiar with the elliptical (circular) motion of the plucked string (Thomas Murach). But I would interpret Elaine Fuller’s (and Simon Martyn-Ellis’) assertion as addressing an optimization.
Think of the fact, that the sound issues from the sound board, which generates the sound waves perceived by the ear, not from the string(s) per se; these are pressure waves generated by the sound board. A horizontal vibration of the sound board is hardly possible, and if it were possible, would not produce any (significant if any) sound. Sound (=pressure) waves emanate from the vertical (up-and-down) motion of the sound board.
Because the deformations in a Harpsichord (vibrating or otherwise) are all so small that linearly elastic analysis is totally applicable. An important consequence of this is, that all motions can be analyzed by addition or subtraction. Thus, the result of horizontal and vertical motions are simply additive. Consequently, the vertical motion of the string is what is most able to excite the sound board. Any horizontal motion of the string does not significantly influence the vertical motion of the sound board; if at all, only through the fact that the string motion occurs at a distance of the bridge height from the sound board, which induces distributed vertical forces (stresses) on the sound board but distributed over probably a larger distance on the sound board than for the vertical string motion, and in addition they are up on one side of the string and down on the other, with the corresponding collective sound board contribution (nearly) cancelling.
I have done a quick, and may be not a truly scientific, test (which any one could reproduce): I have plucked C4 from the keyboard (not really only a “vertical” pluck, I know), and then plucked the same string horizontally, (a) with a plectrum in a tongue, and (b) with the tip of a nail file. In my case there was a distinctly less pronounced sound response for both of the latter cases than for the initial keyboard pluck. For now, I feel satisfied that my reasoning above is also rational, as it also applies to the observation by Elaine Fuller for the lute.
Regarding Domenico’s question of whether this all has to do with the ”bloom” effect (if I understand that term properly), I do not think that this has anything to do with the “wobbling”: rather, the “blooming” results from the fact that the sound has to build over time. The sound cannot be “there” as the result of a single, short vibration, e.g., there is simply not enough energy from a single (or few) vibration(s). I had previously posted a message to this forum in which I showed the growth of the volume of a plucking sound with time. I can probably send that again if you, Domenico, want me to.
Wolfgang