Dave Rifkins article is typical of the sclerotic issues around combining domestic standards and boat standards and as a ABYC certified, he's arguing against himself.
In order to justify the continuous ground argument , he manufacturers a failure mode and then argues for its protection, ie the short between secondary and shield.
Furthermore his assumptions are based on US practice of bonded underwater to AC earth. A practice that is not followed in say Europe
His solution conveniently ignores using RCD devices which on the output of the transformer will provide protection from an obscure secondary to shield fault.
European practice is to use RCDs , not bond AC earth to DC negative , not bond underwater fittings together on GRP boats , where isolating traffos are used the boat AC protective never connects to shore earth
Furthermore Dave Rifken
Would you intentionally go around your own home and selectively sever the green grounding connections on receptacles and applianc- es? Of course not! The same argument applies when the grounding connection is severed on a boat using an Isolation Transformer System!
Perhaps he might vist a UK construction site , where 110 vac isolating transformers are mandatory , the earth is not continuous , or perhaps he's never heard of double isolated tools etc.
His article seeks to take one case and generalise. Every system has fault potential, you cannot eliminate them
For example , as a alternative , run Rifkens case , where the shore AC protective earth , goes hot. Now his solution deliberately compromises the security
provided by the transformer. Every system can be faulted.
A proper high quality transformer , should provide complete isolation , AC protective earth should be local to its own generator
. It should not be connected through to the shore earth ( this is ABYC code) . This allows the transformer to do its job , a d as nick says provide a safety system bar none