Tough question but a good one. Also one that I'm hesitant to give advice specific on since everyone's missions and opinions on redundancy requirements vary. Our generic advice is to understand the instrument/data sources in your aircraft fully, and then you can decide what level of redundancy you need. For example, most piston singles have just one pitot/static system, but airliners have 3 or 4. This is one of the things that gives transport category increased redundancy, at the cost of expense and complexity. In contrast, most IFR piston singles have just one heated pitot, which means that an electrical failure of that single part means no more airspeed, and partical panel becomes your gyros + engine speed + VSI other "informational only" deviceslike GPSes ground speed. They also usually spend less time in severe conditions, like icing, even when IFR.
I'd say 9X percent of our IFR customers running dual Dynons (with at least one battery backed up, of course) run a single pitot/static system with the pitot heated. There are some "really hard" IFR guys that have gone with dual pitots/static system, with one of them plumbed to each unit/ADAHRS.
So, in conclusion, as you're planning your panel, think about what the partial panel is when EACH component part fails in some way - whether it's s electrical, environmental, mechanical, etc. Then, realize that you're building an aircraft for YOUR mission, and that the level of redundancy that you require might be closer to the typical piston single or your neighborhood airliner.
Our generic recommendation - that you'll find through out threads here - is that that you should equip your aircraft such that no single failure causes an inability for you to get back down safely. You'll find that if you have a dual Dynon, battery backed, up, with GPS failure (so that airspeed going away isn't a critical event for attitude), and a heated pitot, you can start "fail" basically any single system in the aircraft and still have flight instruments that can get you down.