The same program was subsequently (in 2001) reissued by Naxos, in transfers by the audio wizard who'd earn the first Nobel Prize of transfers if the category was created, Mark Obert-Thorn, so depending on price you might want to consider that publication: Prokofiev/Khachaturian. It is the one I have, and the audio restoration is great indeed.
I bought it in the context of an thorough comparative listening of Khachaturian's Concerto, and I've published a detailed review under the Naxos entry, which I will summarize here for those intererested in the composition (obviously I can't comment on the quality of RCA's transfers). With Prokofiev's 3rd, Khachaturian's Concerto was a signature piece for Kapell, and this recording with Koussevitzky established something of an interpretive standard - and is still today considered such, sixty+ years after its first publication in 1946 (it was recorded on April 16 of that year. The excellent Naxos liner notes by Jonathan Summers mention that Kapell and Koussevitzky had made an earlier recording, on January 1st of the previous year, but that for reasons unknown it was never released).
And there is indeed, even in view of all the competition that came after, much that is dazzling in this reading, but also a few disputable interpretive options. The most blatant one comes immediately at the beginning, in the way Koussevitzky and Kapell take the opening "allegro maestoso" - at circa 100-104 quarter-notes/minute, much slower that the composer's metronome mark of 108-120, thus giving precedence to the "maestoso" over the "allegro". I'm no radical in abiding by the scores, as long as the interpretive liberties "work". Here, the approach is powerful and majestic indeed, but also dangerously trudging and bombastic, especially when heard against the fiery and headlong dash - at 138, much faster than the composer's indication - of Oborin (the composition's dedicatee and first performer) and Mravinsky (in a live recording made in Prag also in 1946, unfortunately ruled out by dismal sound, Aram Khachaturian: Concertos for Violin & Piano), and of the obscure Czech team of Antonin Jemelik and Alois Klima, recorded in 1960 (Khachaturian: Piano Concerto ; Borkovec :Piano Concerto No 2). They show the benefits of dropping the "maestoso" altogether in favor of a "feroce": gone trudging bombast, enters hair-raising intensity.
That Kapell is capable of exactly that kind of electricity is shown later in the movement, when come the faster passages, which he dashes through at the speed of light: it is electric. Likewise in his finale, which he takes, now (like Jemelik and Klima), much faster than the composer's metronome, 138 against 120-126 - and so much for the better: it is irresistible. And that such electricity is produced at the expense of careful observance of the composer's tempo relationships is of no great matter, since it is so effective that way.
The comment can be extended to the other interpretive liberties taken by Kapell and Koussevitzky, like the adoption in the second movement of a tempo markedly SLOWER than the one indicated by the composer (circa 56 quarter-notes to Khachaturian's 69-72), turning it in the process from an "andante con anima" to a dreamy adagio, or again the phrasing of the first movement's second, folksy theme at 2:36: e.g. markedly slower than the movement's opening tempo (whatever that was), and with no tempo coherence whatsoever when the theme reappers later in the movement and, in slightly modified form, in the second movement. In these various areas, Kapell-Koussevitzky seem to have established something of an interpretive tradition as well: almost every subsequent versions I've heard follows the same options. And (other than the trudging opening tempo) they are effective musically. Nobody without a score is likely to be shocked (and even with it). But the composer's own studio recording with Oborin (sadly, not reissued on CD, but available on U-tube) shows that HIS ways are equally, if not more effective.
Finally, another very disputatble decision of Koussevitzky and RCA is to have omitted the famous flexatone (a kind of musical saw) in the second movement. Was it considered too kitsch, or was there simply none available in Boston in 1946? Anyway, its absence is regrettable. Its kitsch is integral to the music's character. Astoundingly, the composer does the same.
So, yes, this recording certainly has set, in many areas, an interpretive standard. But to claim that it is the only possible view is pure mis-representation. Fortunately, in music, no one interpretation can ever be considered "definitive", and this one, for all its worth, is too far from the composer's intentions and realization to qualify. It is a fascinating view, one everybody genuinely interested in the Concerto should hear, but only one view, and a slanted one at that. In no case can it be taken as the "unique" standard. So if you want a complete understanding of the Concerto, DO "waste your time and money" on more.
(Did I say I was going to summarize? Yeah, well, this is written in August, so I guess you can say I'm summer-izing.)