If everyone followed this rule, there would be very little innovation, none of it disruptive.
If an insurance company specialising in unconventional insurance don't think they can make money out of a premium, that does actually say something pretty insightful about the risk you're taking.
If you'd like a specific example, consider that insurance companies in 48 states can and will likely cancel insurance policies of anyone who rents their car to another individual, even if they hold a separate insurance policy to cover it[1]. This means that the tens of millions of cars sitting in garages would never be shared even if a startup decided to insure those doing peer to peer sharing.
GetAround, which won the TechCrunch Disrupt audience prize last year, or RelayRides which does the same thing, would never have started if they followed this rule. It's a bit early to claim that they have changed the world for the better, but I am optimistic they will.
I think what I said may be somewhat pithy but I certainly wasn't being glib.
[1] http://travel.usatoday.com/news/2011-06-17-personal-car-shar...
More to the point, neither GetAround or AirBnB are in any way deep into territory that insurance hasn't ventured before. Insuring against burglary/damage in rental properties happens all the time, just like for rental cars.
The only times when insurance is unattainable is when the risks highly outweigh the premiums (SEE: Flood insurance in the Mississippi Delta) or when there is no actuarial data available which is tough to come by at this point.
For the first point I'm going to say the parent is very correct in avoiding this type of un-insurable behavior. And as for the latter, you had better have deep, deep pockets because if you can't even convince an insurance company of the risk, are you sure you've thought it through?
I know of few disruptive markets for which actuarial data sets are available.
Obviously you should think through the risks of starting a business not endorsed by insurance companies. If you have an appetite for such risk and believe the rewards (to yourself and/or to society) outweigh the risk, you should do it. Do you think Google could have bought insurance for their self driving cars before they started working on them?