This bug will affects all IP3K products.

When any of the IP3K products is MAC'd it is placed into Factory Test Mode.  If the IP3K device has a wireless card attached it requires that an accesses point with an SSID of "factorytest" be present before it will enter factory test mode.
Once the IP3K device has entered factory test mode then an Ethernet command is sent to the device with the MAC address, UUID, and BoardRev. The values are written into SNV. The device is reset and the MAC address and UUID are read from SNV on the device and verified.

When two MAC stations are used; station one begins a MAC session follow by the second station by about 10-20 seconds.  The first station properly scans and displays the correct MAC address scanned.  The second station properly scans and display the correct MAC address scanned while the first Boom enters factory test mode (looking for access point factorytest.) The second Boom enters factory test mode (looking for access point factorytest.) The first Boom writes the MAC address to SNV followed by the second Boom. The first Boom is rebooted and the MAC address read from SNV and it reports the MAC address of the second Boom.

The only connection between the two MAC stations is with the Booms wireless connection to factorytest. 
Q: Do the two Booms a hard coded wireless address while looking for the access point factorytest?
Q: Could the first IP3K device be "updating" its MAC address in RAM from what it sees from the access point before is it written to SNV?

This is an issue for all IP3K MAC programming devices (Boom, Squeezebox, and Ray) This would make it extremely difficult to run more that one MAC station at any one time.