Well in reality you could take A-E back to the original population since it is suppose to be heterozygous. A person could do the same recurrent selection with THHZ and cross into the THHZ heterozygous line and see which one gets the best results.
So in reality you have a testing line sort of.
I suppose if we wanted to cross it into a deep chunk that say has been selected using the same simple recurrent selection process and produce some potent F1’s and then do an F2 with these seeds (Haze x Deep Chunk:Recurrently selected prior to this cross) and look for the transgressive segregation. Wow full circle again 
.