I have been watching this thread for 2 days and thought I would add my two cents. The transducer switch has been used for years by Humminbird (and most other manufacturers on their product) and I have never seen it to be an issue, before.
As most of you know, when 987 owners had problems reading bottom at high speeds with the SI transducer, the solution was to run a splitter switch box with a dual beam (read more hydrodynamic shaped) transducer on one side of the splitter for high speed running using only the downward reading channel (which gives you your digital depth and your 2D view) and then switch over to the regular SI transducer for slow speeds. It solved the high speed erractic depth problem and worked just fine. I ran one on my boat for two years with no issues what so ever. The only negative was that this setup left the two SI transmitter channels with an open circuit on the output with no transducer or load attached when running at high speeds.
I have seen a few failures on the 997 transmitter transformers that I am fairly sure are all directly related to using the switch box (the last 3 for sure, as I asked). I haven't seen a singe transmitter failure on a 987 or a 1197. That's not to say that there is anything wrong with the 997 transmitter as I believe all the failures I have seen are attributed to the use of the switch box.
I was always taught that you don't run any kind of an amplifier without a load on it. You wouldn't crank your stereo wide open with out the speakers attached and I don't believe it is a good idea to do it with your sonar. The worse part of it all is the SI transmitters are open when your outboard motor is running the fastest and the charging voltage is higher hence a higher supply voltage to the 997 and a higher boost voltage to the transmitter. I believe the transformers are arcing internally, much the same as a failure on a car ignition coil's secondary (high voltage ) circuit.
It really is a no brainer, the Y cable is a lot cheaper then the switch box and now that I am using the Y cable I am not fumbling around looking for the splitter box switch once I come up on plane because I forgot to switch it over for high speed running. I haven't heard of the Y cable causing interference on trolling motors, as RzorBackAngler mentioned, although it could of been one that failed. I had one come across my bench that was arcing internally so badly that I could hear it on my PC's speakers every time I wiggled the cable just below the connector.
I'm not even sure if the Switch Box was a Humminbird solution as I never heard or saw any directives from them to go this route to solve the high speed issue. I heard about it on the forums, as I was checking to see if anyone else was having trouble getting the SI transducer to work at speed with out having the transducer well down below the bottom of the hull which created that great rooster tail we all have known to love. It used a standard available switch that was around since the Wide Side days and I am sure that some rigger came up with the idea and as it worked so well with the 987, everyone use the setup.
Hope this helps and explains a bit