First, a belated reply to the "I think that means that there are no more remote surprises left" - I think there probably are significant surprises left, but they will be connected to entire settlements or urban sites rather than individual monuments (take the Magazberd city site for example, as a published example), and also to the way sites relate to each other.
I finally got round to buying "Monuments arméniens de Haute-Arménie", but I have to say I found it disappointing. I’m not sure how much textual abridgment there is compared to the original articles, but the illustrations in this book are far inferior: photographs are small, dark, and monochrome; the architectural plans are so tiny they are insulting; maps are crudely drawn and sometimes nearly illegible (as if photocopies of photocopies of photocopies). Monuments not visited by Thierry but whose existence is known through pre-genocide texts are also mentioned, and listed individually. While this is useful it also confuses the purpose of the book - is it a detailed architectural description and analysis of surviving sites or a catalogue of everything that is known to have once existed. The book tries to do both. There is also the question about whether the medieval Armenian term “Bardzr Hayk” has relevance to the study or categorization of Armenian architecture.