Understanding output of planemo workflow_lint

It seems that workflow linting currently always warns about at least one `output without a label` for any workflow generated through the GUI editor.
This is because from the GUI you can set a label for input datasets, but not for the contained "output" the properties of which are inaccessible via the GUI. To make that warning disappear you have to open a text editor and add a label in the workflow outputs section of the input dataset step like so:

```
"0": {
"annotation": "",
"content_id": null,
"errors": null,
"id": 0,
"input_connections": {},
"inputs": [
{
"description": "",
"name": "my input"
}
],
"label": "my input as set in the GUI editor",
"name": "Input dataset",
"outputs": [],
"position": {
"left": 227.3125,
"top": 469.5
},
"tool_id": null,
"tool_state": "{\"optional\": false}",
"tool_version": null,
"type": "data_input",
"uuid": "c33af17b-8945-46f5-bdb2-10911d038db8",
"workflow_outputs": [
{
"label": "this output label can only be added through a text editor",
"output_name": "output",
"uuid": "b5f19eb0-775c-4a02-802a-ef715ce8fcc6"
}
]
```

I'm not sure what to think of this:
- why does an input dataset even define workflow outputs (when the dataset is not produced by the WF)? Is this for technical/formal reasons?
- is it a bug in the graphical editor that it does not let you set an important output label?
- or is it a bug in the workflow_lint subcommand that it warns about an unimportant technical detail?

Any guidance what's best practice here?

Start with the planemo workflow_lint subcommand and the GUI-generated workflow example in the report. Determine whether the missing output label belongs in the graphical editor or whether lint should ignore it; done means the expected behavior is agreed and verified for generated workflows.

Written by the indexing model from the issue text.