I'm curious how much the harness plays into this. I'm somewhat surprised by the gemini and grok results, they seem to have strongly deviated from the original images. I'm thinking maybe the harness has a big effect? It's possible to proxy in different models to claude code, if you're curious you might find it interesting to test!
Harness could be a part of it, but worth noting both the Opus and Gemini 3.7 flash tests were both ran through opencode.
The grok test was ran through the cursor cli agent however.