I hate to sound like a broken record, but it looks like trying to support all these third-party stuff is a major waste of resources, not to mention a futile one. There are miriads of them, you will never get them all and always be lugging behind.
You can get the PB Native to support all of them, both present and future by allowing easy consumption of .NET assemblies, which were designed specifically for those third-party stuff.
For example - all you have to do to convert a datastore to JSON or to XML, to anything else for that matter, is to convert it to a .NET class with some collections. There are widely-used .NET modules whose whole purpose in life is to take these classes and convert them to whatever. Let them do the heavy lifting, many of them are free and almost all very cheap, PB developers will not be burdened by the additional costs.
And it works vice-versa as well. If you could just expose a PB native datastore as a .NET DataTable, then there would be a much wider use of PB .NET assemblies in .NET solutions. It is a pain to write C# or VB.NET code for DB selections and updates, so people would gladly use PB for this purpose.
Yes, I know I sound like a broken record.... Sorry about that