I don't understand why if you've gone through this much trouble, you don't have a blog post or even a medium article to cover your findings. I'd be very curious to see the responses from those authors and other experts in the field about your findings.

There's just something dubious about how it seems like you consistently find mistakes and problems in these papers. I'd be stunned if there was any expert that wasn't aware of the shortcomings of using a kernel that's as small as 3x3.