While the poets working in Irish and Scottish Gaelic courts from medieval times onwards have left us a wealth of manuscripts, there are no such corresponding writings from the harpers who worked alongside them.
Whether or not there were harpers writing down their music, it seems clear that there was a large oral element to the Gaelic harp tradition. And since that tradition completely ended over 100 years ago, and few writings by harpers survive, we are mostly hearing about the tradition second-hand, not directly.
The Gaelic harp tradition was however emphatically not a folk, or low-status music. On the contrary, it was one of the highest-status of all arts. In medieval times, harpers were patronised by kings; by the 17th century they were retained in Gaelic and Anglicised great houses. By the late 18th century they had slipped down the social ladder somewhat, but this should not distract us from their former status.
Of course, some Gaelic harpers were able to read music, especially those who worked in courts abroad, playing in consort with viols or lutes. But those playing Gaelic music would have little need to - the music was available much more readily and completely in oral form from other harpers.
So what implications do the above comments have for us, literate 21st century musicians and readers, trying to make sense of the old music? Firstly we might be aware that, as in contemporary traditional music, each performance is unique. Although tunes and formal sets of variations, whether anonymous or by named composers, can be stable and memorised, ornament and variation should probably not be worked out in advance and committed to memory, but improvised or composed on the spot in performance. I also suggest that basses or harmony, to some extent follow the same pattern - far from working out and memorising an "arrangement" of a tune, we might try simply to "play the tune", adding appropriate bass notes and ornament as best fit the tune, tradition and that particular audience's expectation.