feat(core): add configurable QWEN high resolution images support

[AdityaNarayan05]

• Add MIDSCENE_USE_QWEN_HIGH_RES environment variable for controlling vl_high_resolution_images parameter
• Support three states: enabled (true), disabled (false), unspecified (defaults to enabled)
• Only affects QWEN models (qwen-vl, qwen3-vl), ignored for other model types
• Add comprehensive test coverage for all configuration scenarios
• Implement case-insensitive parsing for extended truthy/falsy values

[netlify]

✅ Deploy Preview for midscene ready!

Name	Link
🔨 Latest commit	1e06256
🔍 Latest deploy log	https://app.netlify.com/projects/midscene/deploys/6918c06f28c18a00081d6a92
😎 Deploy Preview	https://deploy-preview-1460--midscene.netlify.app
📱 Preview on mobile	Toggle QR Code... Use your smartphone camera to open QR code link.

---

To edit notification comments on pull requests, go to your Netlify project configuration.

[CLAassistant]

All committers have signed the CLA.

[yuyutaotao]

Hi @AdityaNarayan05 , In which cases should you remove the resolution parameter?

Actually in the coming v1.0 version, we will remove this for qwen3-vl. But it's necessary for qwen2.5-vl according to the document of alibaba cloud.

[AdityaNarayan05]

Hi @yuyutaotao,

Thank you for the feedback! You raise an excellent point about the different QWEN
model versions.

Current use cases for removing the resolution parameter:

1. Performance optimization - When high resolution processing isn't needed for the
   specific use case
2. Cost reduction - High resolution processing may incur higher API costs
3. Compatibility issues - Some deployment environments might have issues with the
   parameter
4. User preference - Allow users to opt-out if they prefer faster processing over
   higher accuracy

Regarding v1.0 and model-specific behavior:
Since you mentioned that qwen3-vl will have this removed in v1.0 but qwen2.5-vl still
needs it, should we update the implementation to be more model-specific?

Proposed approach:
Instead of applying this to all QWEN models, we could:

1. Only apply the parameter to qwen2.5-vl (and other models that need it)
2. Exclude qwen3-vl from this configuration
3. Make it future-proof for v1.0 changes

Would you like me to update the implementation to be more granular based on specific
model versions? I can modify the code to check for exact model names rather than just
the qwen prefix.

Let me know your preference and I'll make the necessary adjustments!