If you were your insurer, wouldn't you want a survey occasionally, on a boat that hadn't (apparently) ever had a proper survey? Just because you assume your boat is in good condition, doesn't mean it would be a good bet for an insurance company, unsurveyed.
I think it's important not to conflate the discussion of liability insurance with hull insurance. For liability insurance, the value of the boat is almost irrelevant and the condition isn't much more relevant.
For liability insurance, in certain situations, it
might be appropriate to require a
targeted survey of a boat to assess
specific risks - such as environmental damage from leaks, propane systems, etc. But beyond this, a general marine survey, even a so-called "insurance survey" does very little to inform the insurance company of their
actual risks. Because there’s no regulatory or industry-wide standard, the term “insurance survey” is informal and variable. It is only meaningful if the requesting insurer is specific about what they want to assess (which they rarely are) and the surveyor limits their survey to just those issues (which they rarely do - though most will if asked).
This lack of clarity strongly suggests that insurers are not using granular actuarial data - but are instead relying simply on heuristics or industry norms. It is like avoiding all snakes because some snakes are venomous, rather than being specific about which are and which aren't. True, you won't die of a snake bite if you avoid all snakes ... and that is approach taken by most insurance companies.
Hull insurance is another story. In that case, it is important that the insurer has a pretty complete picture of what it is they are insuring.
I'm tempted to say that liability insurance is more about who you are insuring and hull insurance is more about what you are insuring. But, admittedly, it isn't quite that simple.