The data-reduction of section access is executed during the opening of an application and will be therefore add some overhead times. The larger an application is the more time is needed to perform this step. I never made benchmarks between my apps with and without section access and therefore I don't know which impact it really has.
But the way how a section access is implemented might make a difference (also never measured in my environment). This means by a quite complex datamodel (for example with multiple fact-tables and link-tables) the data-reduction will need more time as by a neat star-scheme model because the section access worked like the selections within the UI whereby the non-associated data aren't out-greyed respectively set to an inactive state else they will be removed.
This means that approaches which optimize your datamodel itself will quite probably minimize the opening time either directly as also the section access overhead.
Further if you apply a quite complex section access with deep hierarchies over multiple fields you should consider to split the application to be able to apply more simplified section access rules.
- Marcus
