Is the use of the term "Predictors" with regard to IVs in a logistic regression model appropriate or should an alternative term, such as correlates, be used instead?
Specifically thinking of a cross-sectional design.
Depends on what one is doing with it. For example. for multinomial logistic regression one might desire to solve for the individual probabilities, with other possabilities being a set of independent binary regressions, estimating the coefficients, or as a latent-variable model.
One could solve for correlations, but I would take that from the logarithm of a power formula ($Y=X_1^a\;X_2^b\;X_3^c\dots$) to do that (Well, sometimes I would, when it's appropriate).