Why is that? The program I use enables me to "unroll" any surface created to 2D. I print it, wrap it on the uncut stay, cut and bob's yer uncle. That will take decidedly less time than bringing a printed part to a fine polish.
The heat, even to only heat up silver will make for treachery. 6 joints. 24 points.
The other item to knock down in my view from here, the stays are of course tapered, the dropout plug won't be without an interference fit.
You could ream the stay to create a straight bore but that will in my view intensify the problem as the tips of the stays will be very thin.
Alternatively, you can depend on the silver to fill a gap, some formulas will do this better than others
I would suggest not to slot the top and bottom of the stay but slot the dropout, that will reduce the fragile-ness of the stay end. you can keep your side to side detail.
I really think prototyping a joint and "brazing" it up will reveal much and cause a rethink here from the original assembly scheme.
A case that the math can be correct, but the physical will push an alternative.