The inspiration can be taken from run_all_scenarios which allows for safe failures while executing a set of scenarios and also provides an overview of the steps which have been successful for every scenario including failed ones

To have a complete picture, copy-pasting the current workaround (many thanks @davide-f for providing a detailed description in this Discord thread):

A temporary workaround is to locally execute the retrievedatabundle rule with python and download the bundle manually while being logged with your gdrive account. The gdrive limitstion is more strict when downloads are from unknown accounts as in the typical workflow run
To do so you can:

conda activate pypsa-earth
cd pypsa-earth/scripts
python retrieve_databundlelight.py

Here the files are being downloaded and keep notice of the bundles that are not downloaded; most likely the cutout only.

Then go into your config.yaml and ensure to switch: enable_retrieve_databundlelight: false. Then now you can manually download the missing bundles from gdrive. You can find the links into the file configs/bunlde_config.yaml.

As you download your desired bundle, you can place it in the respective folder e.g. pypsa-earth/cutouts.