(A) True. Systematic errors are repeatable biases and can be modelled (e.g., temperature/scale corrections) and corrected through appropriate functional models or calibration.
(B) False. Least squares minimizes random errors given the model; it does not eliminate biases that are not modelled—they propagate into the estimates.
(C) True. Good practice is to detect/model/remove systematic effects before or during adjustment; otherwise they bias the results.
(D) False. Gross (blunder) removal does not imply removal of persistent systematic biases; they are different error classes.