CLIPMergeSimple doesn't work, no matter what ratio selected, only clip1 is used.

### Expected Behavior

If I change the ratio in ClipMergeSimple then both clip models should be merged, but only clip1 is used, even when ratio is set to 1. (From the Wiki: "A ratio of 1.0 means fully adopting the second model's features, while 0.0 retains only the first model's features.")

### Actual Behavior

Clip2 gets ignored and only Clip1 gets used, no matter what ratio selected.

### Steps to Reproduce

Take a SD3Medium Workflow, connect DualClipLoader with clip_g.safetensors and clip_l.safetensors to CLipMergeSimple "clip1" connector. Connect another ClipLoader with t5xxl_fp16.safetensors to ClipMergeSimple "clip2". Sample an image with ratio set to 0.0/0.5/1.0. All three results are identical.
Connect alternately the DualClipLoader and/or the single ClipLoader directly to the prompts to verify, that there is indeed a big difference between these clip models, rendering different images.

### Debug Logs

```powershell
No errors/warnings shown.
```

### Other

Installed (but not used here) custom nodes: ComfyUI-GGUF, ComfyUI_bitsandbytes_NF4, SD-Latent-Interposer, facerestore_cf
Only tested with SD3_medium checkpoint (https://huggingface.co/stabilityai/stable-diffusion-3-medium/blob/main/sd3_medium.safetensors).