If you look at the Decca tree as a three to two vector summing system it is easy to see that by delaying the signals of a three-in-a-line the end result will not be the same as the real Decca tree.
By vector summing I mean a situation where L and M are summed, say, 1:1, then a virtual L is created at half way between L and M. Certainly it is not the same as a real mic in the same spot, but basically something like it. By adjusting timing only, not location, the same effect can not be created, only an approximation. If it sounds better, though, why not use it?
I have made Decca tree contraptions from carbon fiber tubes (about 16 and 14mm, fitting nicely inside each other). T-midsection from thicker tube with stand attachment spigot and then 2 sets of thinner tubes with epoxied screws at the ends, lengths 120 and 90 cm and the front pointing one with one screw 60 and 45 cm (working distance, actual length is a bit less, as the spigot is glued about 5 cm in front of the junction for balance reasons). These are taped in place with black gorilla tape. I can use the side tubes also alone for longer AB miking. These are light, easy to take apart, quite cheap and only one stand is needed.