It's doing what I expected: the "installed" Win 7 on IDE2 Partition 4 boots OK, as does the XP installation on IDE1 Partition 1. The cloned Win7 on SATA1 Partition 2 wasn't "installed" and is almost certainly missing MBR.
There are Linux distros on IDE2 Partitions 1, 2 and 3 and they all boot fine - except that Win7 zapped the Grub configuration that I was using till then for multi-boot. I suspect that a Grub boot would be able to launch the Win 7 on SATA1 Partition 2. I'll dig out a Grub boot floppy some time and try that.
I don't want to zap the Win 7 clone: the whole point of the experiment was to see if I could clone a new, bootable disk easily, from scratch. If I install Win 7 fresh on the SATA drive, I'm sure the installer will tell me it's too late for that (the Beta test is closed to new entries), and also it'll disable one of the other boot options (either the Win 7 on IDE2, or the XP on IDE1).