Thank you for continuing this interesting debate! Yes, I will definitely try to take a good spread of photos for you in the coming growing season. I think you will see that ours is extremely similar to, and possibly not visibly distinguishable from, your repeating version. However, the Trieste plant you have posted clearly does not even have R. setigera in its background and looks nothing like 'Baltimore Belle' (it looks more like something based in R. sempervirens, more like 'Felicite et Perpetue'.) The rugose, "bramble-like" foliage of R. setigera is very distinctive, and of course it is the reason for the synonym "Rosa rubifolia"; I grow and am very familiar with the species itself (which is native to the U.S., of course.) It is normally possible to see the influence of R. setigera's foliage in its descendants through at least several generations of breeding.

The American 'Baltimore Belle' is strongly fragrant. I would describe the fragrance as roughly same as other early noisette varieties like 'Champneys' Pink Cluster' or 'Blush Noisette', but lacking the spicy musk note from R. moschata (coming from the stamens rather than the petals, and only usually in the morning) that some of the early noisettes still possess.

As for the genetics of repeat bloom, typically, when you cross a once-blooming species with a repeat-bloomer, the genetic factor that prevents repeat bloom is dominant in the seedling (which would typically be heterozygous for repeat bloom). In some roses resulting from such a cross, especially diploids, "sporting" to a repeat-blooming form is a well-documented phenomenon. It may not be so much a change in the genes present as it is a change in their expression, or some similar phenomenon; the gene or genes that normally enforce once-blooming may become silenced or partially silenced. This has occurred so frequently in roses and is so widely accepted that the idea doesn't need to be defended here--it happens, that is clear, whether or not we understand precisely what the underlying molecular mechanisms are in each case. However, propagating such a "sport" will often maintain its differentiating characteristics in a stable manner, leading to a new cultivar, while propagating the original will maintain its characteristics. Their DNA may be indistinguishable but they behave differently because of changes in gene expression.

With respect to Buist's comment about the possibility of repeat bloom in R. setigera descendants, you need to understand the full context for that sentence and what he actually meant. Speaking directly about Feast's new varieties ('Baltimore Belle', 'Pink Perpetual'/'Perpetual Michigan', etc.), he says the following: "These roses [of Feast's] will form parents to be impregnated with the more fragrant blooming sorts, such as Bourbon, Tea, Bengal and Noisette. We may therefore expect from them a progeny perfectly hardy, and blooming at least three or four times during the season." The key here is that he is talking about using Feast's hybrids, and not merely the species itself, to cross with the repeat-blooming classes to produce a repeat-blooming generation of seedlings. He is absolutely correct about this; while Feast's roses either didn't repeat or did so reluctantly (see Buist's comments about 'Perpetual Michigan'/'Pink Perpetual'), based on their background they should at least carry the gene for repeat bloom. In the next generation of crosses with repeat-blooming varieties, it would be perfectly normal to expect a certain percentage of repeat-blooming offspring. Even without a detailed understanding of genetics, that much was already understood by some early rose breeders. At the end of Buist's R. setigera section (near the bottom of page 29) he then transitions to his next topic with this sentence: "Having briefly disposed of the tribes of Climbing Roses that bloom only once in the season, a few hints on their general culture will be in place."