Because I2S separates the clocks and data it is (on one hand) easy to copy and (on the other hand) hard to copy because no one makes an easily available off the shelf copying solution.
So, imagine the scenario where we provide standard S/PDIF output (or HDMI unencoded) of the DSD layer. There are dozens of easily obtainable recording solutions one need only purchase and plug in.
There’s nothing we’re aware of that can do this with I2S.
That said, could an engineer figure out a way to build a small FPGA and sort it all out and record it? Sure, but then, that’s not something we’re responsible for because that could have always happened. (It wouldn’t take but a few days and an experienced engineer to reverse engineer the DMP and do the same thing).
As a company we can’t be held responsible for what engineers do to reverse engineer a product. We can be held liable if we make it bonehead easy for a consumer to cobble together off the shelf solutions to violate copyright laws.