This is a tricky one.
Firstly, I would say that if you want to use an HRM for races its doubly important that you have an accurate idea of your max, otherwise, possibly the case here, you are in danger of running too slowly.
Seondly, and this is a personal view, it depends where your lactate threshhold (LT) is. Running tempo sessions in training will help you find out (and improve) your LT but simply speaking if you try and race above your LT, whatever your heart rate might be telling you, then you will slow down.
Thirdly, there is a concept known as "cardiac drift" which estimates that your HR will increase by about 5% (5-10 beats) during the course of a race (as you get tired) for no increase in speed. Thus whilst your percentages might be right for the start of a race, they are probably too low at the end i.e. at the end of a 10k you can expect to be closer to 95-100% of max.
The final factor is more subjective and depends on your own motivation to push harder.
Having said that I have started using an HRM during races this year and found it very helpful for not going off too fast but also even pacing (by building into my race plan an estimated increase throughout the race). Overall though I would say you are better off using a few races to simply gather data (i.e. run as you feel) or experiment e.g. why not try running the first half of a 10k according to your HR estimate, then run the second half without looking at the HRM but run as fast as you can then see what happens to your average in the second half of the race.
