Proposal (CLI): Make defaults more dynamic

We currently have a fixed set of defaults which are set to have a simple effortless experience with a few simple documents.
The defaults are fixed for a better reproducibility of the results.

However making the choice a bit more dynamic could provide better result and overall experience in more cases.

Example of dynamic behavior:

1. The choice of default OCR. If ocrmac is installed, use it as default.
2. The choice of using embedded images in the exported output. If the document is > 20 pages, we could switch to placeholder as default.

All of this should be clearly visible in the processing logs.

Start by tracing the CLI's current fixed defaults for OCR and embedded images, then inspect how processing logs are produced. Define dynamic selection for installed ocrmac and for documents over 20 pages, with each choice clearly reported in the logs.

Written by the indexing model from the issue text.