Pretty cool. I find the approach of using a single memory location to govern the expression of a complex, underlying, much longer "subroutine" sequence very interesting in an evolutionary sense in that it illuminates a mechanism for how and why evolution would layer, nest and control genomic functionality and building blocks.
I often ponder about the evolvability of the DB DNA - the degree to which the form of the DNA itself encourages or discourages the evolution of complexity; such things as the minimum number of base pairs required to perform some specific behaviour and whether evolution can "get there from here" - whether there is a "smooth" path of benificial, incremental mutations that could lead to that sequence which selection can favor over time. The answer to this is critical as just because something is expressable in DB DNA does not necesarily mean that it can evolve. There may be no incremental evolutionary path through the phenotype space leading to that complex sequence. But assuming there is a path, once evolution stumbles on a long, perhaps fragile but highly useful sequnece, controling the expression of that sequence through a single control location makes a lot of sense. That sequence can become stable and much less subject to mutation - a building block upon which richer capabilities can evolve and depend.
In biology, there are incredibly stable genes that go back billions of years that almost all biological organimsms share that have not changed much in all that time. Yet the degree to which these underlying "subroutine" genes are expressed or controlled in different organisms can vary greatly through control mechanisms such as Hox genes. I just find it interesting that we have stumbled upon the similar mechainism of mutating and evolving a single high level expression value rather than the comples, fragile sequence itself.