Really, not many rules regarding deadheading.
Some roses, such as the triploid hybrid musk roses are infertile and just keep blooming and blooming because they cannot form seeds. Some roses, such as rugosas seem to be programmed to continue blooming even as seeds form. Gallicas don't repeat. Period. Ditto multifloras. Most Damasks don't either. Several China roses repeat quite convincingly.
Many floribundas, I would guess, have been selected for quick repeat flowering cycles. And with the number of blossoms a good floribunda can produce it would be silly to require deadheading.
I'm betting that if a diligent researcher were to google all the rose patent applications she would find some mention of roses that repeat willingly without deadheading as a claim in a patent application for a floribunda or a hybrid tea rose. I would guess, too, that it's probably a few decades back (maybe as early as the sixties). If the guess is right then we would expect to have quite a few roses that repeat without deadheading. If this claim were absent from all rose patents, then it would prove that deadheading never actually gets you very much in terms of repeat flower production.
I deadhead informally because it neatens up the plant a little to do so. At least on some plants. But I, too, grow very few hybrid tea roses.
