I am in the process of doing it now. Realized time investment can vary according to how extreme the changes are desired relative to what you started with. What you started with might repressent one breed, many breeds or simply a melting pot of genetic diversity. Not only are you trying to breed toward a given set of characteristics, you are also trying to reduce the amount of variation with respect to performance / appearance. Breeding strategies can be varied even within a given project. I currently use a combination of family selection and inbreeding but will eventually engage in linebreeding as desired product is approached. When trying to fix (make homozygous for) a given trait, then inbreeding is fastest but can result in loss of other desirable traits. Line breeding enables conservation of what looks like a good mix of traits. Family selection can do both desired components of inbreeding and linebreeding plus enable mixing and comparing of different mixtures of traits but rate of change is slower. To do it sustainably, a good number of birds is required, more than typicall for operations maintaining a breed as is.
I use measure of bird generations rather than years. Usually same with chickens until you start getting into business of linebreeding where generations do not mean as much. On paper my project will take at least 20, probably 30 years.
