Is this appropriate to extend for zoomify image slicing?

I like your write-up, and it seems like a good approach to create an actions pipeline to process 'media' files.
I'm wondering if I should look into hooking into your process for the following case:

- I'm setting up an archive/museum exhibit type site (intranet in the first case)
- I have a large number of extremely high-res (> 100MB each) TIFF files of scans of historical photographs to be indexed.
- We are intending to use the zoomify module and scripts to present the images in both preview and 100% resolution. Zoomify works a bit like google maps - many tiles of an image glued together as needed at different zooms.
- To pre-process these images, we need to run a background process to slice super-high-res images into manageable scaled tiles. I'm currently planning how this should be handled.

The original images live on a remote file-store, and will be placed there by the scanners. They cannot be handled by the web interface.

I'm imagining
... a 'harvest' that scans the remote dir for new images.
... a 'process' that imports and slices the big images into the zoomify tiles
... a 'storage' that creates a placeholder resource-type content page where the rest of the metadata will be added to it.

So. Is this process a useful use-case-scenario that goes alongside what you are doing with this module? I see you've concentrated on videos so far, but I can see how my scan-directory harvest process would be a useful addition even to that.
And I see how your ffmpeg external process could have some parallels with my needed slicer script.

I've not examined your code/module yet, I'm just scoping still.
Do you have or plan support for some of the big background processes to be run by cron hooks?
I'm still deciding between D5 and D6 for this. Does/will your pipeline have any cross-over with the d6 'actions'? ... which I currently know nothing about but seem interesting.

ideas?