re 2000 vs 1000, I have never seen anything definitive. However there is a simple, and likely rationale.
For the level cruise case data is based on flight test. For a book aircraft and appropriate met conditions, the aircraft could be expected to make published OEI level plus a delta reflecting the certification fudge factor .. ie the operation is somewhat predictable. In the real world, naturally enough, we put in some fat for mum and the kids to account for age and met conditions even if this is only done on a seat of the pants basis to some extent.
For the drift down case, the certification flight test data may well be repeatable for book aircraft and appropriate met conditions but there is a considerable credibility gap in the accuracy of matching the critical obstacle to the physical drift down locus in 3D space. Remember that these rules go back a long way when the pilot might well have only an approximate idea of his real location. Someone, somewhere in the regulatory business had to draw a line in the sand and than line was 2000ft. Keep in mind also that there is a certification fudge factor in the performance so the book figures provide an additional margin for book aircraft and appropriate met conditions.