Currently, the AF Server will produce identical parameterized SQL queries for 2 AF Searches that search for the same object type with the same filters.
On a large system, performing a search that produces a small result set can cause SQL Server to cache a query plan that is only suitable for small result sets. This can cause performance degradation if larger searches are performed by the same user with the same filter set.
An implementation to avoid this Parameter Sniffing problem for AF Searches could make searching the system more consistent in terms of performance.