Had a convestion with the phys this evening and understand my test data better now.
I cross checked the new tables with my "output" from Abingdon mara and noticed for the first time that the zone I was running at (by intuition 'cos I knew no better?) was entirely in the Zone 4 of the test Lactate Threshold range, so 137 bpm for 1st 10k then, 136, 140, 144 and 141 for the last bit to finish. Had I gone much above 144bpm earlier in the race (per the phys) I would have struggled at the end.
The lactate mMol's were 3.8, 3.7,4.4,5.8 and 4.7. Apparently 5.8 is unsustainable (for me/my level of fitness) which was why I slowed marginally for the last mile or so. NB (note to self: nothing to do with being knackered at that point!)
So what next: if I improve my base endurance at lower pacing, then the whole graph should move to the right and I should access more of the Zone 4 (LT) speed which he estimates goes from 8:24m/mile up to 6:54 m/mile.
For Abingdon (with no science!) I managed 7:45m/mile pacing so I am 50 secs a mile faster than the floor (8:24) and 39 secs below the ceiling (6:54). I haven't got the time to completely close that gap (or maybe even the ability) but if I got halfway then I might be able to shave 20 sec/mile from my overall time which puts me in 3:16:xx territory.
So the new mantra is the same as the old one: Train slow to race fast, except this time I know for certain what "slow" really is, and it sure as hell is a lot slower than I thought it was!