This would work for a short network of less than a couple of meters but for anything longer or with more than two nodes it would allow reflections on the bus that could compromise data integrity.
when a signal transition reaches the end of the line it will reflect if the termination does not match the characteristic impedance of the line (120 ohm).
The reflection will then travel back down the line to the other end. If the other end of the line is terminated with 60 ohm there will be a smaller reflection there as the reflection amplitude is determined by the difference between the line impedance (120) and the termination (60 ohm in this case).
The reflection will bounce back and forth until it is absorbed by losses and the (incorrect termination). All this time though the receivers will get incorrect signal.
If the bus is very short (eg the time of travel is less than the transition time) then the reflections will not have significant effect and the system will act normally. So for testing on a bench with a moderate speed network it can work with both terminations at the same end, but for longer networks it is necessary to follow the termination rules.
CAN Physical Layer Requirements