There is a strong legitimacy to playing Suites made of the purely orchestral numbers from Rameau's operas, the dance movements and other orchestral transitions and interludes, and especially with Les Indes Galantes, the opera-ballet composed in 1735: Rameau himself published (shortly after the first run of the opera production) not the complete score of Les Indes Galantes but (in view, said Rameau in the preface, of the public's dislike of its "scenes", e.g. the non-dance, non-orchestral passages) an arrangement in the form of four suites or "Concerts", arranged not in the order in which they appeared in the opera but reshuffled according to tonality. In truth, Rameau's Concerts include much more than just purely orchestral numbers, but also arias and choruses, and the complete act "Les Sauvages", a later addition to the opera-ballet and not yet represented when the score was published; also, his publication is not that of a version for orchestra, but a transcription for keyboard - on which a number of recordings of Les Indes Galantes with harpsichord have been based, like Christophe Rousset's recent one, Rameau: Les Indes Galantes.
And there is, of course, an even stronger legitimacy, one that overrides any other consideration: sheer pleasure. Listening to the orchestral suites from Rameau's operas is like listening to Wagner's "Ring without Words" arrangements: a way and occasion to revel in the shimmering beauties of the orchestra, without the words and singing getting in the way. Especially welcome for listeners who find that Wagner singers scream, or those, like me, who rapidly become edgy hearing the typical and cliched semi-tone upward or downward appogiaturas ending all the phrases in French baroque opera. The orchestral suites were my entry into the music of Rameau. Recently I heard substantial excerpts from Hippolyte & Aricie - not the suite, the opera itself, and I loved it, so now I'm probably ready for the real thing, but those Suites were a great preparation. Rameau's orchestral idiom is lush, magnificent, imaginative, uplifiting, both highly learned (like Bach's) but drawing on the popular and vernacular, without the profound spirituality that imbues every note written by Bach, but easily as sparkling and festal as Haendel's. When Hippolyte & Aricie made a splash at the Paris Opera in 1733 and revealed the then 50-year-old composer (leaves us all hope, doesn't it?), Campra - possibly in a double-tongued fashion - made the comment that "there is enough music in this opera to make ten of them". Reminds you of what Emperor Joseph II said to Mozart: "too many notes, Mozart!". But it was true, and the comment applies as much to the orchestral music of Les Indes Galantes.
The first recording on period-instruments of an orchestral suite drawn from Les Indes Galantes dates as far back as 1967, by Collegium Aureum, Rameau: Les Indes Galantes; Dardanus [Germany], but stylistically it now belongs to the infancy of period-practice, making the music sound stately, solemn, elegant in a prim and trim manner but also vaguely boring. In 1983, a few years into the big Rameau wave, which began at the end of the 1970s, Philippe Herreweghe recorded a substantial selection for Harmonia Mundi, Rameau: Les Indes Galantes. This one, by Frans Brüggen, was recorded in October 1992 (but published only in 1994), after his previous offerings of Suites from Dardanus and Les Boréades in 1986 (Rameau: Les Boreades Dardanus (Suites)) and Castor & Pollux in 1989 (Rameau: Castor Et Pollux Suite / Purcell: 3 Fantasias). Finally, I've located another version on a small Belgian label, Eufoda, from 2001, not listed here, by Dirk Vermeulen connducing the Prima La Musica ensemble, and one from 2009 by The Aulos Ensemble, Jean-Phillipe Rameau: Suites from Les Fêtes D'Hébé & Les Indes Galantes - apparently reduced to one-a-part - on Centaur. I don't have the two latter.
Herreweghe's and Brüggen's selections don't entirely overlap. Only 15 numbers (out of Herreweghe's 24 and Brüggen's 22) are shared. Two seem to be common, but in fact are not: Herreweghe's "Orage - Air pour Borée et la Rose" (track 17) vs Brüggen's "Air pour Borée et la Rose" (track 20) and the "Air pour Zéphire" (Herreweghe track 18, Brüggen track 19). The opera-ballet had a complex history, wasn't immediately successful, with the music generally praised but the libretto called inept, and Rameau soon made substantial changes, adding a full act (called "Entrées" in opera-ballets), "The Savages", and re-writing about half of "The Flowers" or Persian act. Herreweghe plays the original version of the storm and "Air for Borée and the Rose" (Borée is the allegorical figure of the wind), Brüggen the revision. As for the Air for Zephire, there were two in the original version (and only the first was later retained), Brüggen plays the first, Herreweghe the second. In both cases Brüggen's choices (as were those of Rameau) are better, although it is interesting to have, if only for reference, a recording of the original versions. All these intricate points are made clear, not by the liner notes of any of the two recordings, but by the preface and score of the admirable "urtext" edition published as far back as 1902 by Editions Durand, based on the original manuscripts then kept at the library of the Paris Opera and prepared by no less than Paul Dukas. The International Music Scores Library Project has made those scores, as well as a facsimile of the original edition of the "Quatre Concerts" and other sources, available online, eternal gratitude to them - cash donations won't harm, either.
What remains in proper to Brüggen are the "Air vif" (track 4) and the "Musette en rondeau" from the Prologue (two marvelously boisterous numbers), the "Ritournelle for the Generous Turk" (track 10) and "Forlane of the Sailors" (track 11) from the Turkish act and the "Ritournelle for the Incas from Peru" (track 13).
And for Herreweghe : the Air for the standard-bearing warriors, Air for love and Rigaudons from the Prologue, Loure en rondeau from the Second Entry-the Incas, two Gavottes and an Air from the Flower act.
An additional point in favor of Brüggen is that his concoction is arranged roughly in the order of the opera, clearly distinguishing what belongs to each Entrées, rather than arbitrarily and all shuffled around like Herreweghe's. But the main consideration in favor of Brüggen is that his Rameau is much better played than Herreweghe's, with more beef, bite, sweep and excitement. Although they don't play exactly the same version, Brüggen's storm before the Air for Borée is more stormy than Herreweghe's, the famous Air des Sauvages considerably more savage with Brüggen, Herreweghe's sauvages sound like those Kanaks of the 1931 Paris Colonial Exhibition gathered into human zoos and asked to pretend being savages: they seem to be only half-heartedly going through the motions. You need to go to the track listing of Herreweghe to understand that the "Amants" in Brüggen's track 6 are in fact following Bellone the Goddes of war and that their "Amantes" are trying to hold them back, but with Brüggen they are certainly off to war, with Herreweghe (also track 6) they are simply pretending. Herreweghe's Chaconne sounds impossibly solemn and laden, Brüggen's sparkles and explodes like the concluding fireworks at some royal celebration in Versailles. Brüggen's more prominent flutes and oboes also provide more color.
So, if you are constrained by space and financial limitations, Brüggen must definitely be your choice. I'm keeping the Herreweghe only for archival reference, especially for those numbers not on the Brüggen CD. Among these, I would lament only the absence of the "Air for the standard-bearers" (although it is plain, even without comparison, that Herreweghe plays it with too much solemnity and not enough sparkle), the (second) Air for Zéphire, and the merry Rigaudon(s). Too bad really that these two CDs are so stingy in timing (and by some miraculous chance process, they both have exactly the same, to the second): 43:44. Brüggen would have had ample space to add the Herreweghe numbers and more. At least Universal Europe has recently reissued Brüggen's Indes Galantes-Suite complemented with the Castor & Pollux-Suite he recorded in 1996 (originally paired with three Fantasias of Purcell, see link above). It isn't listed here but is available on the European sister companies under ASIN B004IXJF1K.