Must all science, by definition, always be evaluated as correctly performed or arrive at correct conclusions to come under the definition of doing science?
If by “correct” you mean “accurate” then yes. These are requirements of science.
The hypothesis must be logically grounded in observation.
The experimental design or data to be collected must relate to the hypothesis.
The experiment or data collected must be accurate and reproducible.
The analysis of the data must be accurate with no errors of logic.
The conclusions reached must be accurately based on the data with no errors of logic.
All the above must be open and shown in detail to all who want to duplicate the results.
Absent any of these there is no science.