It's not hard to see what certification costs, take an EAB kit, use all non-certified avionics, and even take an automotive engine and convert it for aircraft use if you want. If you spec that out equivalently* to a certified aircraft, taking labor costs into account (you don't get to count the 1500 hours in your garage as free), it's less than comparable certified aircraft, but not orders of magnitude less.
(By equivalent, I mean safety and redundancy -- ifr capable. Not talking about making my own AP with cheap servos and a raspberrypi and navigating with a handheld gps)
Ultimately, how many new engines are getting certified to put the current ones out of business through lower prices?
There are still many other things in certificated aircraft that are changing regularly based on ADs. This means “we found something in this decades old design that shows up when the fleet is in its 30s.”
Aircraft have VERY long lifetimes in comparison to most things. I own a 1962 Cessna - it is on its second engine, the first having been overhauled once. Some Cessnas from the 70s are still running on their originally installed engine.
Defects and issues are still addressed ongoing, and maintaining the certification for things like modifications for improved power generation still means significant work. It’s definitely not a one-and-done thing - part of the value of certification is in the continuous future improvement that comes from NTSB findings or inspection reports sent to the FAA.
It’s like an engine fault light spread across the whole fleet. The certification keeps the light working.