[Bugfix] Fix request batching regression on Ming-flash-omni ImageGen

[yuanheng-zhao]

PLEASE FILL IN THE PR DESCRIPTION HERE.

Purpose

Fix regression imported in #4079
Resolves #4839

After some bridge func runs, diffusion_prompt is a list[dict]. In vllm_omni/engine/orchestrator.py, Both of the two submit paths reject a list for diffusion stages, however, the unwrap diffusion_prompt = diffusion_prompt[0] get covered only when already_submitted is true.

This PR fixed the above regression, as well as a trivial DiffusionOutput issue in Ming Image-gen pipeline .

Test Plan

vllm serve Jonathan1909/Ming-flash-omni-2.0 --omni \
    --deploy-config vllm_omni/deploy/ming_flash_omni_image.yaml \
    --port 8091

Send a single request with

curl -s http://127.0.0.1:8091/v1/chat/completions \
    -H "Content-Type: application/json" \
    -d '{
      "model": "Jonathan1909/Ming-flash-omni-2.0",
      "messages": [{"role": "user", "content": "Please draw a cute cat."}],
      "modalities": ["image"]
    }' | jq -r '.choices[0].message.content[0].image_url.url | split(",")[1]' | base64 -d > ming_imagegen.png

vLLM Version: 0.24.0

vLLM-Omni Commit: 5c51782

Test Result

BEFORE SUBMITTING: read CONTRIBUTING.md and run the precheck-pr skill with the code agent for a self-check against project conventions.
(anything written below this line will be removed by GitHub Actions)

[chatgpt-codex-connector]

Codex usage limits have been reached for code reviews. Please check with the admins of this repo to increase the limits by adding credits.
Credits must be used to enable repository wide code reviews.

[yuanheng-zhao]

PTAL @yJader , @ZhengWG , @princepride
cc @tzhouam

[yJader]

LGTM

[yJader]

already_submitted is only used here to distinguish the orchestrator submission path (submit_update vs submit_initial). For diffusion, both paths eventually enter the same diffusion client API and then DiffusionEngine.step/step_streaming -> add_request, so DiffusionEngine does not attach different semantics to initial vs update submissions.

Since diffusion StagePool rejects list prompts in both submit_initial and submit_update, singleton-list normalization should not depend on already_submitted. Applying it consistently makes the diffusion payload contract clearer and avoids the initial path still passing a list-shaped payload.

[Gaohan123]

LGTM. Thanks