aggregate table not used in query

tldr

aggregate tables are not being used when time slice is added to query

Want to be able to slice NetSales by day, restaurant, mealperiod, hour, 15min interval and I cannot get the aggregate table to be used when I add an attribute from TIME. It seems like I'm on the cusp, and am hoping someone out there can point out my misstep. I understand why the things I've tried do not work (they should not) but am uncertain where to go from here.

The TIME dimension is

• dual
• TimeKey int 0-1440 for all mins in a day
• Interval15MinKey int 1-96; 15 timekey per 1 interval15minkey
• HourKey int 0-23; 60 timekey per hourkey; 4 interval15minkey per hourkey
• MealPeriodKey int 1-3; many hour key per 1 mealperiodkey

each of the AGG tables are configured in PBI Desktop on top of SalesLabor_BASE. I've confirmed query folding is happening for each. The Agg tables use incremental refresh.

1. SalesLabor_BASE: DirectQuery; SmartDateKey, RestaurantKey, OrderTimeKey, MealPeriodKey, HourKey, Interval15MinKey
2. SalesLaborDay_AGG: Import; SmartDateKey, RestaurantKey
3. SalesLaborMealPeriod_AGG: Import; SmartDateKey, RestaurantKey, MealPeriodKey
4. SalesLaborHour_AGG: Import; SmartDateKey, RestaurantKey, HourKey
5. SalesLaborHour_AGG: Import; SmartDateKey, RestaurantKey, Interval15MinKey

the following relationships are in the model

tried making TIME the base for an aggregate table TimeMealPeriod - no dice

this is the measure I am testing with
MEASURE _Sales[NetSales] = sum(SalesLabor_BASE[NetSales])

from optimizer I can see that the aggregate table is used when the measure, or the measure sliced by date, is called

EVALUATE SUMMARIZECOLUMNS("NetSales", [NetSales])

{
"table": "SalesLabor_BASE",
"matchingResult": "matchFound",
"mapping": {
"table": "SalesLaborDay_AGG"
},
"dataRequest": [
{
"aggregation": "Sum",
"table": "SalesLabor_BASE",
"column": "NetSales",
"mapping": {
"table": "SalesLaborDay_AGG",
"column": "NetSales"
}
}
]
}

EVALUATE SUMMARIZECOLUMNS('DATE'[Date], "NetSales", [NetSales])

{
"table": "SalesLabor_BASE",
"matchingResult": "matchFound",
"mapping": {
"table": "SalesLaborDay_AGG"
},
"dataRequest": [
{
"aggregation": "Sum",
"table": "SalesLabor_BASE",
"column": "NetSales",
"mapping": {
"table": "SalesLaborDay_AGG",
"column": "NetSales"
}
},
{
"table": "SalesLabor_BASE",
"column": "SmartDateKey",
"mapping": {
"table": "SalesLaborDay_AGG",
"column": "SmartDateKey"
}
}
]
}

Here's where the behavior is not as intended

EVALUATE SUMMARIZECOLUMNS('DATE'[Date], 'TIME'[MealPeriod], "NetSales", [NetSales])

{
"table": "SalesLabor_BASE",
"matchingResult": "attemptFailed",
"failureReasons": [
{
"reason": "semijoin not compatible with Import",
"datasource": "Import"
},
{
"alternateSource": "SalesLaborDay_AGG",
"reason": "data source mapping not found",
"datasource": "Import"
},
{
"alternateSource": "SalesLaborMealPeriod_AGG",
"reason": "data source mapping not found",
"datasource": "Import"
},
{
"alternateSource": "SalesLaborHour_AGG",
"reason": "data source mapping not found",
"datasource": "Import"
},
{
"alternateSource": "SalesLabor15_AGG",
"reason": "data source mapping not found",
"datasource": "Import"
}
],
"dataRequest": [
{
"aggregation": "Sum",
"table": "SalesLabor_BASE",
"column": "NetSales"
},
{
"table": "SalesLabor_BASE",
"column": "SmartDateKey"
},
{
"table": "SalesLabor_BASE",
"column": "OrderTimeKey"
}
]
}