The OP's second try doesn't look too bad. Now I do think that since very few people try to stitch anything themselves - that is why so many are poo-pooing the idea. The way I see it, arborists are constantly doing things themselves that if done improperly, result in a fall, death etc. Just the way it is. I am a rigging engineer who frequents nuclear, DOD and DOE jobsites. We take rigging procedures and requirements to the extreme. On some lifts we have single-failure-proof cranes and rigging or else go to 10:1 rigging safety factors. No one goes under a load let alone reach under one. All sorts of stuff. But I go home and can be hanging from a single part of 7/16" rope. A single carabiner, single pulley or many other possibilities. Often one failure away from falling on my head. Certainly makes you think twice until you get used to it and/or start coming up with ways to add redundancy.
If I ever get around to playing with sewn splices, I'll dissect a known good splice, count stitches, measure thread dia. etc. Watch videos, do some reading, buy the best thread there is then give it a try. Doesn't mean I'll go around selling spliced gear but my stitching is going to be the least of my worries. That said, I'd probably rig up to a couple trees and do my own pull test to see if I need to dial in my process. I have a 10,000# digital load cell and plenty of real rigging to break a small rope and not get whipped. In rigging we use off-the-shelf equipment when available. If not then we do our homework and design up what we need. Fabricate, load test, go to work.