PandaBear, engineering and actuarial work are surprisingly parallel in a lot of ways, and ways that I've seen.
Neither are a pure science.
Quite often engineering jobs are unique, but that doesn't mean that they are trial and error, they just take some known values and experience, do some extrapolation and risk analysis, and most of the time get it right.
Actuaries, as I've found out over the last 20 years, do exactly the same work, but backwards, using known failure rates of known components, operations and loadings, applying even more conservative safety factors, and apportioning a dollar value to the risk.
Offering the customer a reduced premium, provided that they reduce the risk, in the areas that they feel uncomfortable with, and negotiating a premium/risk profile that either enables the job to proceed or kills it stone motherless dead.
As soon as you say "I'm not sure" to an insurance guy, you will be finding out for yourself, and quickly. If you say "I don't know", you'll be battling for credibility to even be able to negotiate a policy.
And before someone jumps in and mentions "market at work"...it needs some regulation as to responsibility, and quite possibly a punitive component to work as such.