Generally, I prepare solutions of the substances to be tested at various concentrations in the reaction buffer, then add the enzyme and substrates to start the catalytic reaction.

I don't usually do a preincubation. A preincubation would be desirable if you are trying to reconstitute the enzyme with a structural metal ion (commonly zinc). If the metal ion is a co-substrate, it should reach equilibrium with the enzyme very rapidly, so no preincubation would be necessary.

Since metals and salts are normally added in large excess over the enzyme under catalytic conditions, their free concentrations are not significantly affected by binding to the enzyme.

It is always good to pre-incubate the enzyme with the additive of your interest (metal ion, salt etc.) keeping all the parameters constant (enzyme conc., buffer, pH etc.) and varying the concentration of the additive. Generally, you will get an sigmoidal curve with a sharp change after a certain concentration (dose response).

Proteins need free water as "grease" for conformational changes. Added osmolytes (salts, sugar, glycerol...) will reduce the available [water] and thereby limit motion in the protein. This will increase stability and crystallisability (that's why proteins are crystallised from osmolyte solutions), but decrease enzymatic activity.

Pre-incubation is used when binding of a factor requires time (for example, I used it for measuring the effect of a neutralising antibody on enzymatic activity), if binding is near-instantaneous (diffusion controlled), pre-incubation is not required and may be detrimental if the enzyme has limited stability.