Step 1: Properties of orthogonal matrices.
A matrix \( P \) is orthogonal if \( P^T P = I_3 \), meaning that \( P \) preserves the length of vectors. This implies that all eigenvalues of \( P \) lie on the unit circle in the complex plane, i.e., they have an absolute value of 1. Since \( \det(P) = 1 \), the product of the eigenvalues is 1.
Step 2: Analyzing the options.
Option (A): A matrix \( P \in \mathcal{O} \) with \( \lambda = \frac{1}{2} \) as an eigenvalue is impossible because the eigenvalues must have an absolute value of 1.
Option (B): A matrix \( P \in \mathcal{O} \) with \( \lambda = 2 \) as an eigenvalue is also impossible for the same reason.
Option (C): If \( \lambda \) is the only real eigenvalue of \( P \in \mathcal{O} \), the other two eigenvalues must be complex conjugates, and the product of all three eigenvalues must be 1. Hence, \( \lambda = 1 \).
Option (D): Since \( \det(P) = 1 \) and \( P \) is orthogonal, it is possible for \( \lambda = -1 \) to be an eigenvalue (this would correspond to a reflection matrix). Thus, the correct answer is \( \boxed{(C)} \) and \( \boxed{(D)} \).