You are right that performance data will come from the manufacturer's flight test data. However, it is usually spot-checked (at the very least) by the certification authority during type certification. Each individual production series aircraft must meet the spec performance for its initial and subsequent C of A airtests.
Airspeed indication errors may be due to a variety of causes. The obvious starting point is the indicator itself. Or is there a problem with the pitot (dynamic pressure) sensing probe alignment? (Possible blanking.) How easy is balance keeping? An aircraft with poor out-of-balance tactile cues could result in the airflow into the pitot head being reduced by inherent yaw. A static blockage or partial blockage might affect things, but you might also see effects on the other pressure instruments.
Commonly, hanging bits on aircraft may upset the airflow past the static ports, so this is often investigated when modifications are being approved. Flights over an accurate course in opposite directions, using raw timing, GPS, kine-theodolites, a suitably calibrated 'chase' aircraft or a combination of these methods is used to extract data to compile or update pressure-error correction charts. (This should be legislated as part of the modification submission and the design organisation would, in UK, have to have prior approval from the CAA flight department for the pilot actually flying the test.)
There are some answers.
If you have serious doubts as to the accuracy of indicated airspeed you should bring this to the attention of your maintenance organisation.
Now would you like to indicate the aircraft type, please?