The catalog tables are used to determine the access plan chosen by DB2 to process the SQL. For example, DB2 will look at the various indexes and their statistics as stored in the catalog to determine whether to use any of them instead of doing a table scan. However, keep in mind that if the same exact SQL is being executed more than once, the access plan "may" be cached in package cache, or if static SQL is used (such as found in most stored procedures) the access plan is already stored in with the package.
So yes, reorg and runstats of the catalog tables can help with performance.