Thanks for the additional layer. I have just evaluated the predicition over the last few thermal flying days. It is always important to note that thermal predicitions are also a bit regionally influenced, so my assessment is mainly valid for the thermally weaker area in NW Germany.
What I notice is:
The thermal height predicition is usually overoptimistic. Often 200m below is about right. And I am already talking about the best height achieved by the best pilot of the day.
The thermal max. height is an important parameter, but the thermal avg. strength would actually be much more useful and better to compare.
a) To begin with, the thermal height can be easily derived from the temp, which an advanced user will anyway look at. Thermal strength is less easy to derive without actual calculations and it is not so that high base = strong day.
b) Over mountainous terrain always the highest peak have the best thermal max. height. They do not always have the strongest thermals though, usually the better ones, but not always the strongest. Thus the prediciton becomes a terrain map there rather than a thermal forecast there.
Hard to say why the forecast is nearly always overoptimistic in my region. One reason might be that ground dryness plays in reality a role in how good the ground warms up and that is perhaps not used in the calculations.
Also anything a bit closer to the see gets see wind effects, though not in all wind situations and I find the predictions also over-optimistic in south wind situations.