Adobe certainly has a lot of manpower, but they won't fix your bugs and release an update the same day it was reported or even write down your feature requests. 🙂
The main reason we don't license the source is the risk it puts on our business. This is separate from my opinion that getting involved with setting up and forking a complex tool is not usually necessary.
It depends on your use case as to how much of a hack it is. The various textures need to be very similar to be applied to the same mesh, so all that is involved is a visual check for replacement textures. It's not uncommon to have a tool for previewing skeleton configurations and associating/editing game data. This could allow artists to preview their mesh texture swapping, hot reload images, etc. The skeleton viewer tool is nearly what you'd want for this and it's only 150 LOC (without UI).
There is a similar situation which is common, where an app has hundreds or thousands of images per slot. It doesn't make sense to do the work of attaching them all in Spine just to preview that the art is correct. It's much more efficient to setup only a single image per slot, like a template, then programmatically replace the attachments with the actual images needed. In this case you'd also want a tool artists can use for previewing.
Best of luck figuring out how you will be building you game! If you have any more questions or concerns about how Spine might work for you, feel free to ask them here.
