Thank you for all these responses. Very interesting discussions. I'm using the mileage & elevation just for personal records and data for planning future hikes.
I understand about small differences in elevation but when one system shows me 770' for total gain and another shows me nearly double that gain for the same hike, I'd like to know which is closer to reality. I could get a detailed topo, trace a route over each bump & saddle to get an estimate. Does anyone know how the apps are getting their data since they don't ask for stride length they must be only using GPS (Lat/Lon) but then both apps should match each other at least but only sometimes (usually on miles). How often are the apps polling data (every 1 sec? every 10 sec? when detecting motion only? constantly?). I know the Garmin is adjustable and mine is set for 7 seconds currently.
I mostly use AllTrails for looking up possible trails that I may have missed or trails to link together other hikes I've done. But sometimes they show trails that are private but not marked that way in their system. We had an angry rancher threaten us in Chatsworth once even though we were still on a public trail according to AllTrails. But I find the AllTrails interface a bit clunky too.
Strava I like the interface more but to keep a running tally of my mileage & elevation, I have to set my hikes to "Runs" because their software was designed more for cycling and running.
Garmin seems to still be buggy. After a hike, I use Garmin BaseCamp to examine my route. It's a great program for really analyzing sections of hikes or planning future hikes. But that's where I've found issues. That's where I can highlight all of the points on a single track but get sometimes radically different results if I don't highlight the first or last point (doesn't matter which). Sometimes it's not too big a deal like 4.3 miles changes to 4.1 but the longer your hike the more that completely changes the data. I did San Jacinto a couple years ago with the Oregon 650t. My route (which was set to every 20 seconds back then) without the last point showed 10.2 miles & 2600' gain but when you include the last point it jumped to 14.1 miles & 2700' gain. The moving and stopped times also changed by 30 minutes. All of the hiking books show it is much closer to 10.5 than 14 miles. When I tested other hikes against book mileage, the Garmin usually matched closer when you remove the last point. And that hike the other day changed my total time from about 4hrs to 6days+5hrs if I didn't delete the last point.
It would be a good experiment to do those 3 hikes again but I try to never repeat a hike. I suppose I could for the sake of science or just to satisfy my own curiosity. But those three hikes would not be ones I'd choose to do over 2-3 times for data because they involved quite a bit of off-trail, overgrown, tick-infested portions. But maybe I will pick a couple nice trail hikes of varying terrain and test the 3 methods two or three times to see how the results stack up.