Have you evaluated a model for a conventional design where the seat stays meet the seat tube roughly in-line with the top tube? I’d be reluctant to draw conclusions based on the stress you’re seeing in the dropped stay model, but a relative comparison between the two would be useful. If you’re seeing a similar stress in each they’re probably roughly equivalent structurally. Definitely be cognizant of details that aren’t represented in your model like heat input, distortion, post-weld reaming, etc. like @ElysianBikeCo mentions. Based on failures I’ve seen online there’s sometimes not much difference between frames that last a long time and those that fail prematurely - details really matter, even small ones like cable guide placement (though that isn’t especially relevant here).
Also, purely out of curiosity, what is motivating the use of a 35mm seat tube for a (presumably) drop bar bike with a 27.2 seat post? I don’t think I’ve seen that before.