So, after wandering the AOPA Summit, I've heard a few times that the FAA still wants to go ahead with their Chart Tax(or chart user fee).
As the government can't actually copyright things they publish, I expect this will be handled in a 'trade secret' type manner and requiring contracts between each party in the chain from the FAA to the end user.
Obviously this won't work for chartbundle.
The alternative seems to be going back to the good old days and scanning things from the paper copies.
Right now our system shows that the approximate number of 'sides' needed to be scanned for VFR charts every 28 days is about 30. At $10/scan that's about 300/mo. Plus maybe an hour for a human to do checking, so let's call it $30/scan for $900/mo.
For the IFR charts, every 56 days we have 81 new charts(as every chart is updated every 56 days)
Assuming they can be better automated as they're much simpler, let's call them $20/chart(scan+georeference). $1620/2 months(56 days)
The more interesing part is the approach plates. Right now, we have 17000 charts loaded for the last cycle. Assuming you could get scan+ocr+referencing for $0.10 per scan, that's $1700/56 days+some amount every 28 days for the change notices. This would require a ton of automation to check the version number and build reference data so the charts are usable.(Charts that don't change wouldn't need full re-referencing, reducing some costs.)
AFD is about 5000 pages, so $500/56 days.
The chart costs themselves add a few thousand.
Adding this all up, for one company it would be about $42000/yr to do all the charts themselves.
(Assuming wild fudge factors) With in-house scanning hardware the ongoing expenses could be traded for capital expenses.
If one could get 10 organizations sharing costs, it's then $4200/yr. More than that it would be an even better price.
Hopefully we don't need to get to this point, but it's financially feasible if a number of organizations got together.