Answer: No
Reason:
Dale is biased toward the two fastest people in the group, but is leaving out the rest of the swimmers who may not be so fast. The people left out may be very distant from the top two swimmers. Therefore, Dale has not picked a representative sample.
As an example with numbers, let's say we had these average speeds in mph: {1, 2, 3, 4, 20, 30}. Of course these are just hypothetical examples and not based on actual realistic speeds; however, the example should still suffice.
That data set shows the average is closer to 2 and 3 than it is to 20 and 30. Dale polling the fastest people greatly skews the mean larger than it should be.
To get a better more representative sample, Dale should randomly pick 2 people regardless if they are the top two swimmers or not. He can pick the names out of a hat, or use a computer program to generate random numbers.