I have to wade in here! The ferroresonant power supplies are horribly inefficient under light loads. Several years ago we replaced 15 hops of tube type microwave with solid state. Each repeater site had a Sola ferroresonant voltage regulator and they all ran slightly warm. After we reduced the load, the transformers ran so hot they would cause a skin burn and actually burned the paint on several sites. We ended up making a trip to remove every one of them to prevent a shelter fire. The load went down and the circulating currents in the Sola transformers went way up. If you have a power supply feeding a repeater, it is sized for the maximum transmit power demands and during receive time the load goes way down and guess what, the efficiency goes to pot! Ferroresonant transformers can also increase the damage caused by a sharp rise and fall time spike. They can actually ring and cause one spike to turn into a bunch of spikes with lots of energy that stresses components.
As far as overvoltage damage is concerned, that can happen on both analog and switchers, that is why you use a good stiff crowbar circuit to protect the equipment. It is very common to have a "reach through" failure on a series pass transistor, where a hole gets punched between the emitter and collector causing a short. This then puts the full regulator input voltage on the output, yet the emitter base and collector base junctions will test OK! Don't get me wrong, I would much rather work on an analog supply than a switcher! They both have good and bad points, but sometimes the situation drives one into favor over the other. I like switch mode supplies where size and heat are issues and I like analog supplies for the simplicity and lower parts count. Ferroresonant supplies have their place, just not at my place! Jack K6YC