First of all, you didn't sound like you had much experience with underwriting or with hurricanes.
Building a home that is 'guaranteed' to withstand a Cat 3 or 4 storm would make it prohibitively expensive. Forget computer modelling; when winds get past 120 mph there are too many variables based on individual structures and landscaping. Near or above 120 mph wierd things happen.
Although most believe that the coastlines are the worst, there is significant damage miles inland from hurricane winds and tornadoes. Also there are more deaths and damage inland from flooding than there is near the coast.
Regarding damage and risk assessments, one account I worked was about $138 million and based on my report and assessment the insurance carrier still made money after it was hit by Hurricane Ivan. So I have an idea of what goes on.
You are absolutely correct that housing is very difficult to build to withstand a cat 3 or 4 storm. It's rather like trying to "earthquake proof" buildings as Kobe Japan thought they did. I don't live in an area that's subject to hurricanes, and sounds like you do. However, I am very aware of the affect of catastrope losses and/or adverse regulatory/political climate on the insurance industry - fire, hurricane, earthquake, or whatever.
I have underwritten property insurance for about 30 years, personal lines and small commercial, not large commercial, so you've got me there. I've worked for national and regional carriers and large and small independent agencies and been involved in pricing/regulatory issues for the last several.
As you point out, when a major hurricane hits, it can cover a very large area and damage can be/has been in the billions.
I don't know what you mean by the one account, 138 million, making money. If you mean that a single, commercial account had more premium than loss on a particular hurricane, that could well be true. However, pricing is usually set based on multiple accounts - preferably lots of accounts.
In commercial, there is individual risk premium modification (based on aspects/predicted profitability of a particular account) but that occurs by account or risk and after the "basic" pricing is set.
The point I was trying to make is that the risk of loss in Florida for homeowners may be beyond what many preferred risk companies may be willing to gamble, even if they have other lines of insurance there and even if they can raise their rates at will (which they cannot due to regulatory issues). And there may not be enough low-exposure risks in the entire state to adequately subsidize the ones that are most exposed to hurricanes. Last but not least, those folks will be cranky when their rates go up in order for coastal properties to get lower rates.
States not subject to as much severe weather as Florida would also be cranky if a large national carrier like State Farm or Allstate put a "Florida hurricane surcharge", obvious or not-so-obvious, in their other states' pricing.
I have seen carriers pull out of or go broke in Hawaii, California, New Jersey, Texas, Florida and other states when they could not get adequate pricing for the exposure. Availability and affordability, interestingly enough, tend to be reduced when the regulatory climate gets too onerous. "All or none" risks the possibility of none.
It is also notable that insurance companies sometimes have short memories and market share and price cutting become more attractive as the latest disaster gets older.
Insurance is a numbers game, for profit, like any other business. No profit, no insurance.