Most likely the problem was poor quality tube metal or poor quality solder/weld material.
If there is no pencil zinc for the H/E I'd not worry. THAT might offer some protection.
"The mechanic suggested bonding to the H/E to the frame and to the bonding system."
I am unsure what that means: If you connect a wire from the H/E to the frame of that generator you have done all you can. [See below] You do NOT want to run separate wires from the H/E, one to the generator frame and another to the bonding system. Generator ground and any other ground should have a single point of connection to the bonding system. Don't add any others. The problem with an isolated H/E like this is that any protective zincs associated with the bonding system and galvanic corrosion or electrolysis protection are too far away to protect the H/E. That's why for example Detroit water pumps and heat exchangers have local pencil zincs.
I had the H/E apart on my 1972 Hatteras Onan 12 KW generator at age 35...it was starting to run hotter than I liked.... I found the iron ends of the H/E were so overgrown with iron oxide [rust], and maybe mineral deposits from salt water attached as well, it was blocking the water passage through the copper tubes. That H/E is now 42 years old and is still working just fine...no zinc, no bonding....BUT
In all probability your H/E is "bonded" [ Loosely speaking, really grounded] all this means is that it is electrically connected to ground and any bonding system via the mounting bolts. It's easy to check: Test the resistance between any part on the H/E and either the generator or if you prefer either ground or a point in the bonding system. If it IS electrically connected you'll get almost zero resistance way,way less than an ohm....
If you'd like to connect a wire anyway it sure won't hurt anything. You may also find that when you generator is off, salt water drains out of the H/E....if so 99% of the time any electrical condition makes no difference.
On isolation transformers: Isolation transformers keep most dockside frailities and screw ups of other guys off your boat...like reverse polarity at the shorepower pedestal, dc leaks in the boat next door, lack of zincs on the other guys boat, and so forth. If you like everything bonded, do it whether you have an isolation transformer or not. I have posted several times on the pros and cons and most people like to follow ABYC rules to be safe. If anything, an isolation transformer means you'll be likely to use up your own on board zincs more slowely. But you'll still want to use zincs in some places to help guard against stray current corrosion in the outside water, galvanic corrosion due to different underwater metals on you boat, and the possibility of corrosion due to dc leakage currents on your own boat. In other words, if you liked bonding and local zincs without an isolation transformer, you'll still like them with one.
A prime example of where local zincs are required are to protect bronze alloy propellers from corrosion in the presence of stainless steel shafts...so shaft zincs in the vicinity of the props are always a good idea.
PS: I have been surveying an aluminum hull Burger with the owner of my former Hatt. It may be of interest to note that much of what I said above for a non conducting Hatteras fiberglass hull does NOT apply to metal hulls. Aluminum is more electrically active than typical underwater Hatteras metals like bronze or stainless steel. Burger, for example, electrically ISOLATES all through hull seacocks from everything else. The seacocks are all stainless steel and bronze, very corrosion resistant in themselves, are NOT bonded and NOT zinc protected. But zincs on the shafts and pencil zincs in the main engines are deployed as we have discussed.
