Sure! I don't use the GP tool, but you could. More info here: Create Database View in ArcGIS Desktop I usually use the database management tools to prototype the view's SQL Script and then use ArcCatalog to create the actual view, however I've recently revisited this problem and changed my approach entirely.
I have a feature class of addresses, and a related table of what is essentially contact information. The contact data in the related table are updated with 4 options from the web app front end: Contacted, Not Contacted, Attempted, and Other. To symbolize the address layer by what's in the related tables I added the address layer to my Table of Contents 4 times, and used separate definition queries for each one (renaming them to match the query used).
The definition query is just the "WHERE" clause of a SQL expression, which means you can select from other tables in your Enterprise Geodatabase easily. for example my "Not Contacted" layer's definition query looks like this:
EID in (SELECT EID from PS.CANVASS_FIRE_EVW where ACTIVE = 'Yes' and CONTACTED = 3)
Our EID is a GUID, so you could use Esri's Global IDs for this as well (and that would be better, I'm just working with some real legacy data). What's happening here is I am selecting only the features in my address feature class with a unique identifier (EID) that is also in the related table (PS.CANVASS_FIRE), is ACTIVE, and has the CONTACTED status I am looking for (3 in this case is Not Contacted). This nicely gets around Oracle's 1,000 value limit for "in" statements as well by using the related table for the values. If your related table is versioned you need to use the versioned view (indicated with the _evw and created by SDE automatically when you version a table or feature class). This allows your symbology to be updated live as the data is updated, rather than after a compress.