Thank you Geoff for your informative reply. I contacted MM technical support and they told me that the figures are reasonably accurate as the terrain is measured every 50m and interpolated within that distance and when the route goes at right angles to an incline the measurement is accurate but when a steep slope is travelled along the slope "quantisation" produces a much higher reading. That is what you said.
Counting the contour lines does not take into account any ascents or descents less than 10m between those contour lines so the number of contour lines crossed can give a serious understating of ascent/descent.

I'm a bit late on this - been busy!.
The ascent figures from Memory Map (or any other software) are nearly always too high. It's down to the nature of the elevation data:- it is a height grid giving average height in each cell of the grid. The plotted route or track repeatedly cuts across squares of the grid, no matter how accurately you plot it (although the more accurate the plot, the smaller the error).
This is easy to see from the route profile, the plot of a level path will look 'bristly' because of the large number of little ups and downs as it slices through the height grid.

The general rule is:- the closer to perpendicular the route is to the contours, the more accurate the ascent figure, and the closer to parallel, the less accurate it tends to be.

For a kurtose route such as a mountain horseshoe - a good example is the Snowdon horseshoe - the ascent figure is quite good. For a route that goes up the sides of valleys,the figure is too high. The worst case is a route that traverses a steep slope, which gives a figure much too high.

Is anyone able to tell me how accurate are the statistics on MM particularly those relating to ascents and descents?
I recently went on a group walking holiday on Exmoor and the amount of ascent and distance were quoted on the walks info sheet. I took careful note of the routes on my OS map and transferred them to MM when I got home but I found that although the distances compared favourably the amount of ascent was considerably different. These were the figures:
Info sheet Day 1 1950ft MM 3118ft
Day 2 1500ft MM 3035ft
Day 3 2360ft MM 3291ft
It said in the info sheet that "ascent is calculated on the basis of contour lines crossed on the OS maps which is a more reliable indication of height gain than those produced by new technologies" I would be pleased if anyone can explain the wide differences in these figures.
And what system does LDWA use for calculating total ascent on LDP's.
Thank you.