I was fortunate to have assisted DJ in testing several gliders and he/we did measure sink rate by timing descents. Typically in 1,000 ft increments. Also, we/he did several flights for critical data points and threw out data from flights that obviously suffered from atmospheric or data anomalies. Data points from the remaining flights were manually averaged to create the finished polar. And, he did normalize the data to sea level, standard atmosphere, and TAS conditions. There was at least one article in Soaring in which he published the formulas. Interestingly, when I first analyzed data and created graphs using Lotus 1-2-3 (does that age me or what??), Dick suggested I was wasting my time as he trusted calculators and graphs created manually more. I'm pleased to see he eventually accepted Excel outputs.