alright....how i see it.....
let's assume that all your birds, the spangled, and the laced, are eb/eb based (more on this later).
parents then should be:
spangled: eb/eb Db/Db co+/co+ Pg/Pg etc etc (rest the same)
Laced: eb/eb db+/db+ Co/Co Pg/Pg etc etc (rest the same)
so 2 genes really at play, Db, and Co. Db/Db is required for spangling-it effectively takes lacing from around the feathers and moves it to the tips. If you add Co/Co then you move that spangling back to a barred like line across the feather and end up with reduced spangling-and this is where i think you are seeing the spangling fading with subsequent moults, because Co is more of an adult restrictor than a chick restrictor.
okay, so all first generation will be split:
eb/eb Db/db+ Co/co+ Pg/Pg etc for simplicity these are Db/db+ Co/co+ and to get your laced back you need Co/Co and db+/db+; to get spangled back you need co+/co+ Db/Db. these 2 genes do not co-segregate....so each individual gene will make its own effect in a single kid, so you MUST get the combo together again to get the "proper" kids back. Twice as many birds right there need to be hatched.
so now, you cross brother and sister. and sister and brother.
you will (without going into the gory details) get
1/32 kids will be spangled (co+/co+ and Db/Db). 1/32 kids will be laced (Co/Co db+/db+). Then there's a mess. 1/32 will be double laced. 5/32 will be half spangled (and probably hidden as they get older). 2/32 kids will be incomplete spangled (no clue how you will tell them apart from half spangled or double laced/half spangled). 2/32 kids will be incomplete laced (no clue how you will tell these apart from laced, or double laced/half spangled). 4/32 will be double laced/half spangled (again, no clue how you will tell these guys apart from the birds you want).
Now, remember this is just for females or males, you will need 64 birds (statistically speaking) to get 1 male and 1 female of each of your laced and your spangled back. so....if you didn't hatch 64 birds minimum.....especially this year where everything seems to be males LOL
you will ONLY get your laced and your spangled back in your next generation IF you have successfully selected the fully laced and fully spangled birds, otherwise you will be restricting your chances of getting them in future generations but not guaranteeing that you WILL get them. unless you carry a large number of birds and cross breed.
Now, if your dark cornish are like mine and i think they are split e+ or e+/Ewh, or ewh/eb.....this greatly complicates things. because now you have 3 genes you need to sort out into the correct combos to get spangled vs laced. the main drivers are still co/co and Db/Db (with Ml/Ml but cornish carry Ml so it's all good). but the e locus will be mixed. if they are ewh/ewh you then will also have to select based on chick down colour in your second generation-the first generation will all be ewh/eb, but the next is 1/4 ewh/eb, 1/2 ewh/eb and 1/4 eb/eb it's all good if you want eb/eb but ewh will cover eb, so you will have a tough time telling ewh/ewh from ewh/eb chicks. and right there, need another doubling of animal hatched numbers. so instead of 64 kids, you'll need 128 kids to get 1 male and 1 female of each.
I am going to assume though that the cornish on both lines are on the eb locus....
and if they are on the ewh/ewh locus then the same argument as above applies....64 birds minimum.
and this all assumes that the spangling is done through darkbrown and columbian and patterning genes...and not mottled.....LOL. if it's through mottled then Robbie would be correct up there!