Of course, Bob. The more crossovers you have left the smaller the effect of immediate smoothing is; this is pretty obvious. To be honest, the position you posted would hardly qualify as one that needs serious inside distribution improvement, so the rule does not really apply. How about the following variant:
        
 
         
 | | White is Player 2
score: 0
pip: 77 | Unlimited Game
Jacoby Beaver | pip: 90
score: 0
Blue is Player 1 |
|
| XGID=---BBBEAA-AA-------hcbb---:0:0:1:41:0:0:3:0:10 |
| Blue to play 41 |
| 1. | XG Roller++ | 10/5 | eq: -0.864 |
| Player:
Opponent: | 24.51% (G:0.00% B:0.00%)
75.49% (G:0.00% B:0.00%) | |
|
| 2. | XG Roller++ | 8/4 7/6 | eq: -0.866 (-0.002) |
| Player:
Opponent: | 24.48% (G:0.00% B:0.00%)
75.52% (G:0.00% B:0.00%) | |
|
| 3. | XG Roller++ | 8/4 6/5 | eq: -0.866 (-0.002) |
| Player:
Opponent: | 24.48% (G:0.00% B:0.00%)
75.52% (G:0.00% B:0.00%) | |
|
| 4. | XG Roller++ | 8/4 3/2 | eq: -0.868 (-0.004) |
| Player:
Opponent: | 24.44% (G:0.00% B:0.00%)
75.56% (G:0.00% B:0.00%) | |
|
eXtreme Gammon Version: 2.10
I do agree with you that the rule cannot universally be applied to an arbitrary odd number of crossovers left. David never formulated it in such a general way; it clearly works best with one crossover and usually is right on the money with three. I never regretted using it in practice so far. It never hurts to save 0.005 or so of equity now and then.