I think it has at least partially to do with different people measuring vampire lose differently.
Some people charge the battery, disconnect the charger, and immediately check the kWh value. The then check again say 5 hours later and use the difference as their vampire loss. It might be just the batteries settling after the charge. i.e. the reading just after a charge might be a little off. It is not really lost since it was not really there to begin with.
Other people drive for a while, park, then check the reading. When they check check again after 5 hours they are probably getting a close reading to true vampire losses.
Then again there are many other factors, like temperature, use of the App or visible Tesla, use of Dash cam, fobs too close to the car, etc. People rarely list all the details of their set up when reporting the values. That is the problem with Anecdotal evidence. If the tests are not done under controlled repeatable conditions, it's hard to really tell what is going on.