The problem appears to be corruption of the disk's private region:

 A dd of the problem luns: One from the source system and one from the destination system:

 "dd if=/dev/rdsk/  of=/var/tmp/ bs=1024k count=64" 



Source system disk.out looks good:
cat disk.out |od -c |more
0000000   E   M   C   -   S   Y   M   M   E   T   R   I   X   -   5   7
0000020   7   1       c   y   l       3   6   8   2   6       a   l   t
0000040       2       h   d       1   5       s   e   c       1   2   8
0000060  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0

Comparing it to  the same output from a test kit, we should see the following at 0020000

0020000   V   x   V   M   D   I   S   K  \0  \0  \t 316  \0  \0  \0 002

What we see on the BCV clone disk is zeros in that area:

0001000  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0  \0
*
3601000   P   R   I   V   H   E   A   D  \0  \0 031   è  \0 002  \0 002

Volume Manager will not recognize this disk since our header information is corrupted.  

Rerun the BCV establish and split routines. Represent the luns to the second host and verify that the private region info is intact.