This is something I’ve thought about. For instance, Rosa wichuraiana is very resistant to mildew, but ‘Gardenia’ [R. wichuraiana x Perle des Jardins] is susceptible. To be fair, ‘Gardenia’ is not quite as badly afflicted as ‘Perle des Jardins’, but it is not always healthy looking. I have thought about crossing ‘Gardenia’ with ‘Yvonne Rabier’, to get another dose of the Wichuraiana mildew resistance.
It’s possible the original double white he used may have had some fertility, but the ones we have here, now are a degenerative sport which has been propagated forever until that’s all we have.
But there have been exceptions, such as “The Monster”, bred from R. banksiae banksiae x Old Blush. It may be gone, but two of its offspring continue on: ‘Lila Banks’ and ‘Old Lady Gates’.
Roses can vary in their fertility, influenced by soil, situation, climate, and I don’t know what all else. Even back in 1840, Rivers wrote, “Banksian Roses seldom bear seed in this country; but in the South of France, and in Italy, they produce it in tolerable abundance; so that we may yet expect crimson and other coloured roses of this charming family.”
Boursault raised his hybrids in his Paris conservatory. Tunningley, in Schenectady, raised “The Monster” in a greenhouse. Advice to folks who aren’t traveling to Europe anytime soon.
There are other cases of stubborn roses:
Nicolas (1927)
We have long noted here that grafted plants of R. Hugonis, for example, will profusely bear seeds, while plants grown from cuttings are very scant seed bearers, almost approaching sterility. Paul’s Scarlet Climber as an own root plant may be considered as sterile, but a grafted plant will bear both self- and hand-pollinated seeds.
Climatic conditions may also have something to do: a rose hybridizer of experience in Canada writes that Mme. Caroline Testout and Gruss an Teplitz are totally sterile for him, yet they notoriously are prolific seed bearers in other sections and in Europe.
http://bulbnrose.x10.mx/Roses/breeding/Nicolas_sterility.html
Jacobs (1937)
Father Schoener undertook many experiments with those native flowers, but his initial attempts at hybridizing and domesticating them were disheartening. He chose the hardy wild Nutkana as a parent stock because it could withstand the cold northern climate and it bloomed early. He selected the Paul Neyton [sic], an old and large French rose, as its mate. Working literally like a bee, he pollenized 1,500 blooms the first spring. But the only result was to prove what he had learned in his botany classes: that a wild rose will not take pollen from any other species.
Then his next step was to graft the Nutkana onto the vine of the Paul Neyton as a sort of blood transfusion. It worked! — and after fertilizing this plant, he obtained five perfect fruits. The product of this seed is Schoener’s Nutkana, well known to rose lovers — a large, single pink rose, which sends up shoots seven to eight feet high each year and produces bunches of flowers from every eye along the stem.
http://bulbnrose.x10.mx/Roses/breeding/Schoener/SchoenerJacobs1937.html
Van Fleet (1919)
Owing to its poor seeding abilities when grown as grafted plants on heavy soil, less progress has been made than was hoped for with _R. Moyesii, _notable among wild roses for the deep red coloring and waxy texture of its widely expanded blooms. Now that our plants have been transferred to the sandy loam of Bell Experiment Plot, and have become established on their own roots, seeds are more freely borne, and a fair number of hybrids are under way. Pollen was plentifully produced, even when the fruits failed to mature, and a few early crosses, the result of applying it to the stigmas of other species and varieties, have sufficiently developed to show prospective value.
http://bulbnrose.x10.mx/Roses/breeding/VanFleet/VanFleet1919.html
On another tangent, normal plants can sometimes be made abnormal by fiddling with their environmental conditions. I suspect that the opposite — making abnormal plants relatively normal — might also be possible. The following report discusses (among other things) an ordinary that was induced to make five falls by messing with the temperature during storage.
http://bulbnrose.x10.mx/Heredity/thermomorphogenesis/thermomorphogenesis.html