Use User Cost Attribution Data to spot anomalies and make part of our Recommendation Engine

UCA is a great way to report on the ongoing costs of your production AVD environment. However, using it it's also easy to spot anomalies. Example: the FSLogix storage costs are largely equal for most users, however, for some, the costs are much higher. Easy to spot by sorting on costs within the dashboard. Can we use this data, come up with a calculation of some sort where we focus on the (average) costs per configuration item, or make it preconfigurable (FSLogix storage being the example here), and if any of the more important ones stand out, flag them and make this part of the recommendation engine, to further investigate, or come up with an recommendation, preferably?