We see it from two different perspectives, but neither is less valid than the other. Personally, I am able to "have in my head" a harmonic succession without having to play it, and therefore I do not feel the need for it. In fact, I will tell you more: where it proves impossible to play certain passages of some particularly complex scores on the piano due to obvious limitations, the ear can overcome those limits, because it goes far beyond ten fingers... I studied Stravinsky's Petrouchka score at the desk and (technically speaking) I would only know how to play certain passages, but that score is in my head with its exact timbres, its exact tempos, and without any "digital" filter.
The need to explain to another person is an aspect that is not part of the objectives of a score reading course, because if that were the case, they wouldn't ask you to read Mahler's 3rd on the piano, but they would ask you to read "these eight bars of brass," do you see what I mean?
What I find completely useless, frustrating, and absurd is forcing people (even non-pianists, damn it, what did they do wrong?) to sit in front of an orchestral score and read entire pages by performing an extemporaneous synthesis. Synthesis is only possible after analysis, and pianistic synthesis, where possible, would require a knowledge of the subject (meaning, with "subject," the piano) which in itself requires a lot of time. What could a clarinetist (or whoever) gain from that exercise after a couple of years of "complementary piano" when faced with a score?
If I were to draft a program for a score reading/complementary piano course, I would think, above all, about providing non-pianists with a useful tool for their profession; therefore, for example, I would skip Beethoven's Sonatas and put into the repertoire, for each instrumentalist, a selection of solo pieces from their respective repertoire, of which the student in question should learn the pianistic reduction of the orchestral part or the pianistic part in the case of chamber music. Thus, violinists would learn to sketch "Spring," Tchaikovsky's Concerto...; flute players would learn Mozart's Concertos, do you understand? This would have a utility, pianistically speaking, in their daily study.
As for the specific part regarding the score, if one must, I would look favorably upon a targeted exercise on instrument families, with particular reference to transposing instruments, which, when they must be inserted extemporaneously into an overall reading, are ignored in most cases (the high woodwinds, first violins, and cellos-double basses are read, and don't tell me it's not true!). Taking Mahler's 8th and reading two pages of only horns is a truly formative experience; taking the Pictures [of At the Gramophone] and reading Catacombae (just that one, have you ever tried?) without peeking at the pianistic part in the tiny tiny staff below makes you feel really stupid, but that is what would really be useful.
Another thing that would be useful (but here we are getting into complicated territory) would be to get used to singing orchestral parts, something that 90% of conductors cannot do (just like 90% of instrumentalists, who, if you show them a score, have no idea what it is until they have sketched it on their instrument...). But to be able to reproduce a part autonomously, it is necessary to have assimilated it with the inner ear, and we are back to square one.
All this, naturally, is seen from the perspective of an orchestral conductor, but really, just one—me. It is my point of view and I do not pretend at all to claim that it is the only valid one, even because I am refuted by the facts, namely by many colleagues who normally use the piano for studying. I, despite being a pianist, have never done so and continue to feel no need for it...