Insurance for relay failure is getting more granular because relay risk is no longer perceived as one event class.
Underwriters are separating corridor exposure, duty type, recovery depth, and contractual dependence rather than treating every communications lapse as a generic outage. A relay that supports archive sync is not priced like a relay that carries continuous control traffic, and a corridor with layered fallback is not priced like one built around procedural trust alone.
The practical effect is a market that understands infrastructure more intimately. Operators with better routing discipline and stronger reserve paths can increasingly demonstrate why their failures should cost less to insure. Those without that discipline end up revealing their architecture in the premium.
Insurance used to summarize relay risk after the fact. It is now starting to interrogate design before the failure happens.