but this also makes the behavior of SNN models more difficult to interpret #1

When you say this, what do you mean? Typically, the fact that the neurons are now binary makes an SNN easier to interpret as you can look at the activation patterns when doing things like saliency mapping.