Most of this is noise; the idea of opposition between gating networks and basic methods is an oversimplification.
A gating network is just an iterative refinement, a more sophisticated way to incorporate evidence for predicting venture capital success, not a fundamental contradiction.
Basic methods already use textual descriptions and structured data; gating adds a layer of dynamic control for instance-specific weights.
Instead of opposition, it is an evolution in managing multi-source evidence.
For example, integrating regulatory compliance requirements from the FSA via a gating network does not contradict the need for an initial risk assessment but enhances it.