comfyui/README: link the custom node repo, be precise about the diffusers gap

The conditioning tensor is bit-equal to example.py (checked: max abs diff 0.0,
(1, 251, 7680)), but the image is not pixel-identical to diffusers: ComfyUI
supplies its own text-position ids and 512-token padding on top of its sampler.

comfyui/README.md

@@ -4,7 +4,8 @@ This folder contains a ComfyUI workflow that drives **FLUX.2-klein-4B** with the
 **Qwen3-0.6B + adapter** text encoder instead of its native Qwen3-4B encoder.
 
 - Workflow: [`flux2_klein_qwen3_06b_adapter.json`](flux2_klein_qwen3_06b_adapter.json)
-- Custom node (the two `KleinAdapter` nodes): **klein-qwen3-adapter-comfyui** on GitHub.
+- Custom node (the two `KleinAdapter` nodes):
+  [github.com/recoilme/klein-qwen3-adapter-comfyui](https://github.com/recoilme/klein-qwen3-adapter-comfyui)
 
 ## 1. Install the custom node
 
@@ -44,8 +45,10 @@ guidance 4.0.
 
 ## Notes
 
-The node computes the text conditioning byte-identically to `example.py` in this
+The node computes the text conditioning identically to `example.py` in this
 repo (same chat template, same layer taps `2,9,14,18,23,27`, same fp32 adapter,
-same `drop_first=5`). The final image differs from `diffusers` only by framework
-sampling (noise RNG, scheduler) — the same difference you'd see with the native
-klein encoder.
+same `drop_first=5`) — checked numerically, the two tensors are bit-equal
+(`(1, 251, 7680)`, max abs diff 0.0). The image itself is then produced by
+ComfyUI's own sampler (its noise, scheduler, text-position ids and 512-token
+padding of the conditioning), so it is not pixel-identical to `diffusers` — the
+same framework difference you'd see with the native klein encoder.