Doggone it, Dave, how did you reach into my brain and end up saying the same things I was going to post?
What is the most important feature of an aircraft engine? Some will jump up and say, "Power to weight ratio," others will point towards low specific fuel consumption, and many will center on low cost.
But the most important feature of an airplane engine is reliability. Without reliability, any other virtue of an engine is meaningless. It might be capable of producing ten horsepower per pound, burning half a gallon of ethanol-laced gas per hour, and cost only $500 to build, but all this is to no avail if the plane ends up embedded in the trees off the departure end of the runway.
Is it possible to build a reliable auto-engine conversion. Heck yes! Talk to William Wynne, Ben Haas, and others with long-flying aircraft.
But...but...the track record for auto-engine conversions is not that good. In the studies I've made, I estimated that auto-engined homebuilts have a 23% higher accident rate than those powered by traditional engines.
The problem is not the suitability of the engines themselves, but the abilities of the folks converting them. People experienced with engines may not have much trouble, but someone who doesn't understand engines that well can get into trouble. They don't have the right background for how things should be done, and they don't have decades of experience to detect when a problem might be brewing.
Here's a chart from a presentation I gave at my EAA chapter. This compares the number of hours on the aircraft at the time of accident, for traditional engines and auto-engine conversions. Note that over 20% of all auto-engine homebuilt accidents occur in the first ten hours...about three times the rate of conventional engines. This indicates that too many homebuilders can't get their engines running reliably.
The main areas where auto engines suffer is because of those systems that Lycoming/Continental users don't have to contend with: Reduction drives, radiators/water pumps, and electronic ignition. Yes, electronic ignition is more reliable than the old-fashioned magneto. But it is NOT more reliable than TWO old-fashioned magnetos, especially when a single electrical-system failure can kill the engine..
The best auto-engine solution, for a typical homebuilder, is to buy a "firewall forward" conversion from a reputable company. This...surprise, surprise... usually ends up being sold at a significant percentage of the cost of a rebuilt Lycoming/Continental.
I'm all for folks experimenting with alternate engines. However, folks have to understand they ARE experimenting, and that experiments do fail. If you are building to have an aircraft that you can enjoy with the least worries, you're better off with a traditional engine. A very respected local homebuilder installed a Ford V-6 conversion in his BD-4. He was an engineer, with extensive engine-rebuilding experience, using a commercially-available reduction drive system.
What, you're expecting some disastrous finale highlighting the folly of auto-engines in aircraft? Not from me. He installed the Ford after the BD-4's original Lycoming failed on takeoff and he crashed. He flew that airplane with the Ford engine for a number of years, all across the country, even to Oshkosh a couple of times.
But he bought a used Bonanza to use whenever he was taking his family anywhere. He recognized there would always be uncertainty regarding the with the auto conversion's reliability. He was willing to take the risk himself, but wasn't willing to risk his family.