> Does this mean that future hardware releases ie 3.x, 4.x ..., the samples undertaken > will all be compatible with the hardware 3.0?
There was an issue with the analogue filter on the pre-3.0 boards that was causing distortion of the signal due to an unexpected high-frequency harmonic (that wasn't even documented in the Laserdisc specs) - so the main change on 3.0 was a new elliptic filter design that successfully removes the issue.
The 'incompatibility' is really noted for those that don't really understand the underlying process... ld-decode (a separate project that decodes the RF samples back to video/audio etc) needs to be 'tuned' to the response curves of the duplicator. Changing the filter changed the response quite a bit - so, if you use a 2.x board to make the RF sample, it won't necessary give a good result with ld-decode.
The actual sample 'format' is just data - so there is no 'file' incompatibility. Provided there are no issues with the analogue stages of the 3.0 design (and so far all testing has been very positive); there should be no need for such a design change again in the near future.