A requirement is an agreed upon item to be worked ... when an ask isn't inline with functionality of the tool it should ever become a requirement. 

You could do it by setting a session variable, but not a repository variable ... this is how currency selection and exchange rate is done. Here's some example of how it has been done: INDEXCOL( CASE VALUEOF(NQ_SESSION."PREFERRED_CURRENCY") WHEN 'USD' THEN 0 WHEN 'EUR' THEN 1 WHEN 'AUD' THEN 2 END, "01 - Sample App Data…

what you need is a dimension value that represents all organizations and your measure logically aggregated to it OR plot a level-based measure (built in RPD) at Org-Total on a second vertical axis.

https://blogs.oracle.com/pa/entry/obiee_11g_installation_best_practices

what is in the stage table that is missing from the presentation warehouse table? Use drive physical design, so perhaps a second presentation table generated from the stage is required for that use -- I agree not to open up access to your processing layer.

"I try to show counts of employees by age" + "haven't found satisfactory solution with this that works properly for leap years, etc." those statements alone beg a properly formed physical structure. The simple approach is a snapshot fact that stores either the detailed employee & age at a snapshot date or it stores the…

Use of information drives physical design ... why not use the database (and functions you know work) to update the age attribute column in the dimension nightly in your post-load etl processing?

Agreed ... so if the requirements stands then apply the rule: Use of information drive physical model design -- and restructure the data so it's accessible in an easier way to BICS.

Good point it could be spotty in it's application! Best to avoid that setting (explicit deny on auth user) in the first place. 

This is more of a report and not on-screen analysis ... quite frankly a bi dashboard tool is the wrong tool for this job. If you must do it in this tool ... then structure your data physically so you can use the tool. So think about how you could get the data into an HTML output first (leverage the DB to do your work) -…

The BI Administrator Role has an override ... for example this type of thing and row-level security doesn't apply to an admin. An explicit deny to Authenticated User applies to everyone EXCEPT the Admins. 

+1 to @2807661 ... manage the permissions through the catalog.

What you are struggling with is 'query grain' ... really to show the correct information is to split out the agent from the vehicle (via the method @Christian Berg has mentioned): Agent Name | level based measure at Agent | Vehicle | base measure…

So you have mixed logical dimensional attributes? you should have a fact (amount) joined to both a date (year, month, date) and an account (account number, first name, last name, etc) dimension. OBIEE requires at the very least a well-formed logical star model to perform to its advertised features and performance levels.

Your logical date dimension needs a hierarchy where month is keyed logically higher to date ... then BI server will aggregate the measure based on the aggregation rule supplied in the fact logical column and from/to the levels you've specified in the Content tab of the fact logical table source properties. Once it's…

"_SHOW_CORRECT_VALUES=true|false option" is @Gianni Ceresa and/or @Christian Berg Now that you've used it ...

Now that you have a snapshot you can configure the logical table sources and hierarchies to control your summaries (so you won't be adding up snapshot elements over time but using the last aggregation rule and you can add up like summary levels such as Status or Error)

Did you try implementing your logic into a new logical column? post what you did in that column and what you are receiving as 'not working right' or error messages.

+ Giannni's response of getting your own sandbox - then 'development' becomes the integration box

A seasoned modeler wouldn't assume and would go in to each one any way ... providing the BI server with explicit instructions removes a lot of the potential for unexpected behavior.