Hi Gents,
First of all, apologies for my "non-attendance" over the weekend due to other commitments (

okay, so they were only jam sessions - but live music is something of a priority!).
I'm pleased to be able to announce that, after considerable practice and your patient responses, I've managed to locate a fermata on several occasions (perhaps 10% of the time) by ensuring that the bar on either side was empty. I realised on reading pcartwright's and wcreed51's posts, however, that my original objective was somewhat forgotten when posing my initial question. Unable to place a fermata over a
double repeat barline , I'd decided to simplify matters by trying it first over a single barline and my question arose from this. For reference, I am indeed trying to "notate as Brahms did" (or rather as Simrock, his publisher, did) and the score in question is for the "Haydn Variations". Most of the variations consist of two repeat sections and the fermata are placed to indicate a "stylistic pause for breath" before the beginning of the next variation - in that respect, they are pauses between measures.
To return to my progress in solving the present problem, I was at least successful also in placing a fermata over a repeat-start barline but came unstuck when putting a repeat-end barline in the preceding bar - as did the fermata, which was nudged to the left of the repeat-end. Try as I might, I could not get the fermata to stay on top of the barline! In the end, I gave up and decided to listen to the result to make sure that the pause wasn't where I wanted it to be anyway (i.e. that it wasn't just a graphics aberration), so I inserted a 2 beat pause in the fermata and played back the segment. I didn't actually hear the pause at all!
Taking it further, I decided to try the experiment on a violin test part I'd created to check out GPO4 playback. I inserted the fermata over a repeat bar set as before and added a further repeat passage on to the original part. Needless to say, I again failed to hear the pause, so I exported the test file to a MIDI file and viewed the results in Reaper's events editor. As expected, the events list looked like something that should have Bolognese sauce poured on it (Notion's "rules engine"

, the reason for the original GPO test file) but it was when I looked down the events list for the end of the first repeat that I began to question my sanity

! To say that Notion's handling of repeats on MIDI export is "a little lacking" would indeed be the understatement of the year!
Since I need some confirmation that I'm not hallucinating, I've attached my test file herewith so that somebody else can kindly export it to a midi file as I did and compare the events list to the score. Please assist in confirming that I'm not prematurely senile!

(If I'm not, then I'll submit a bug report, otherwise I guess I'll just have to resume the medication

)
Thanks again in anticipation,
Keith