Mark, not as a lurker but as a very nerdy girl (additionally intrigued by yours being post #2010!) I've decided to roll up my sleeves and look into your question.
Let me tell you upfront that I've not been trading this, and that it just seems to me that it would be more profitable to add the extra 2% you're willing to risk to standard NT.
OK, here are the answers (spread in my data is 2pips, so this is MNT32S):
2009 overall: +544
Months:
1 -64
2 108
3 -12
4 38
5 -1
6 80
7 138
8 68
9 38
10 36
11 28
12 87
2010 (up to end of March): + 138
Months:
1 46
2 58
3 34
Standard (I think this is vanilla?) NT:
2009 overall: +1677
2010 (up to end of March): +175
So actually, I stand corrected! Up to march this year, the two were pretty neck-to-neck. In 2009 however, if 2% would have allowed you to bet £5 per pip, risking 4% on std NT would have banked you £16,770, whilst combining NT and MNT 50-50 would have yielded only £11,105 (and even less on combinations giving more weight to MNT of course).
Barbara
PS I should add, the data I have might differ by some pips from the ones from IG etc; as i understand it, each broker or sb firm has its own (hopefully small) differences. NT and its combos are however very sensitive - at times we win or lose by just 1 pip, so this can make a difference for individual trades. Not on average though over such a long period I've looked at for you.
PPS I too would be interested in a private forum.