I think I should provide the catch-22s for self-pollinating daylilies.
Diploids
Most diploid daylilies that have been scientifically investigated are known as self-incompatible. If you self-pollinate most fertile diploid daylilies the flower will act as if the cross took, a pod will start to develop but then after a week or more the pod will abort. Most self-pollinations in diploids will fail to produce good seed. Self-pollinations of some diploid cultivars do work, examples are the lemon lily, Hemerocallis lilioasphodelus or H. flava, Stella De Oro, H. minor, etc.
If a diploid self-pollination succeeds then the second problem rears its head. That is, self-pollinations are inbreeding at the strongest rate and inbreeding causes loss of fitness in daylilies. Many of the seedlings from self-pollinations may be weak, may die, not bloom well, on average fewer seeds may be produced, etc. Selection for good characteristics will tend to be more difficult - that is, on average require more seedlings.
Tetraploids
Making a plant tetraploid sometimes eliminates self-incompatibility and greatly reduces the negative effects of inbreeding. Both of these are probably true for tetraploid daylilies. Tetraploids may be self-compatible more often than diploids - self-pollinations may work and produce good (viable) seed. Since inbreeding occurs at a much slower rate its effects are much smaller and may not even be easily noticeable after one generation of self-pollination. I have been told that one hybridizer had a tetraploid self-pollinated line that was acceptably vigorous after many generations of self-pollinating it.
