Wotcher Lubble
I think we do something of that nature here, but not with UDTs. We have a process which populates/updates PAY_RETRO_ASSIGNMENTS with relevant recalc dates based on changes to things like Grade rate.
With anything like this, the trick is to be able to establish a list of assignments that make use of the entity that has had a change of some kind. In the case of stuff to do with grades, that's fairly straightforward - the grade_id is on the assignment of course.
For UDTs, however, the water becomes a whole lot muddier. For one thing, it's highly likely that you will have many different UDT structures defined which rely on completely different items of information to determine the end result cell value. The issue is further complicated by the number of rows/columns that might be defined. If that wasn't complicated enough for you, you have to add in the fact that ultimately, the reference row/column values that are used to seek a particular cell value might themselves be derived in some way - at the end of the day, UDTs are accessed by formulas (via the get_table_value function of course) of which there may be many, and themselves all different. In short, your custom code would have to pre-empt what cell value the formula would have returned from the relevant UDT, in the case of each assignment
There, have I cheered you up now?