To my mind the evidence, though not conclusive, points the other way.
Though none of the contemporary reviews quoted in Peter John Smith's 2008 thesis "The Choral Music of Charles Villiers Stanford" (available online) state specifically what accompaniment was used for the Gloria, they leave no doubt that the full London Symphony Orchestra was carted to Cambridge to play the works needing orchestral accompaniment, and one article particularly praised Stanford's conducting of the chorus and orchestra in Parry's Blest Pair of Sirens. This is not exactly proof, but is it plausible that Stanford would have let the orchestra sit there doing nothing during his Gloria? Feverishly productive as he still was even in 1920, I should have thought he would have written the orchestral parts out himself rather than waste the occasion.
The idea that the Gloria was sung with only organ accompaniment was suggested in Smith's thesis, but he was under the mistaken impression that Stanford had left the Mass in short score only. In fact, the full score had been sitting in the British Library all the time, but had been temporarily lost sight of - it was part of a collection of MSS deposited there by Boosey & Hawkes but not actually acquired by the BL until a few years ago, following which they were all properly catalogued. Forgive me if I do not look up the dates for this, but the BL holding is easily accessed online and full details are given of its history. If Stanford had really never completed the orchestration of the Mass, then Smith's theory of an organ-only performance in 1920 would have been much more tenable.
The fact that no orchestral parts are known to survive is the one counter-argument to my belief that the 1920 performance was with orchestra. However, considering how much has disappeared - including at least two further Mass settings (a cappella this time) by Stanford, I cannot see this as very conclusive proof.
Lastly, the Gloria was not the only movement to have been aired prior to the Cardiff premiere - the Orlando Singers gave the Kyrie at the Wexford Festival on 10 August 2018. In this case, there is no doubt that organ accompaniment was used.
None of this is intended to belittle the achievement of all concerned in presenting the complete work for the first time at Cardiff and of Lyrita in presenting the first recording of a piece that will hopefully assume its rightful place in the choral repertoire.
Thank you for taking part in the MusicWeb International Forum.
Len Mullenger - Founder of MusicWeb