I feel as though I got a little insight on the Pumpkin genetics when I purchased two Rocky Top Pumpkin Mini-Games from Rocky Top Game Farm in Mississippi. As with Pumpkin Hulseys, some of the offspring come out "white". Well, I bred the "white" one I got to some oversized red quill American Game Bantams and all of the offspring turned out with dun in the black parts. She isn't white, she is a dun/fawn mutant/sport (which ever terminology you care to use). 100% of her offspring were dun where they would have been black. It had an interesting effect on the red quill pattern.
Some have mentioned the blue dilution gene may be at work, but I now believe its the dun/fawn dilution gene. The female I have has similarities to a splash except that the splash marks are smaller and are a salmon orange, not blue. she also has the salmon breast, as would a BBR, but it is very light. For experimental purposes, I plan to do a test breeding between her and a black OEGB that I did one of my first dub jobs on this year that I saved from the cull pile.
My understanding on Serama was that there was already a "butterscotch" in Serama that was essentially the same as the pumpkin in LF Games (American as well as several of the Asian game fowl breeds - unofficially). In fact, I saw one at the Pacific Northwest Poultry Association Winter Classic last year in Stevenson Washington, except his cape and saddle were golden over the butterscotch body. It was quite striking. Unfortunately for his owner, he had a really bad comb. She thought the world of him and he did have great conformation otherwise. Unfortunately, when you have a uneven comb with huge gaps between the points, well, it's the first thing anyone, including a judge, notices regardless of the breed or how they "present" when you toss them on the table. But, I'd be happy if I could recreate the look of that plumage in an American Game Bantam.