I'm not saying he (or you) shouldn't try to estimate BMR. In certain situations it can be helpful. Just that it should be obvious that these calculations are not overly scientific, and shouldn't be thought of as anything more than a very vague concept of energy needs.
For example there is nothing to account for a change in nutrition or calorie status, or for training length and intensity, thyroid output, or any of the other million systems in the body that can alter calorie expenditure.
In OPs case, even though his weight has remained constant, the calculation has led him to deduce that his maintenance calories are around 2300 when they're actually evidently about the 1700 he's presently eating.
The best calculation to use for a BMR is whatever you're eating, under whatever training programme you have, that maintains body mass at approximately the same level for a period of at least 3 or more weeks...