My spontaneous instinct would be to go with the third option and read out the locations on the map, convert it and add an icon on the minimap.
You could add a capsule to your character which has a radius as large as the area the camera captures. Then you get all your overlapping Location_BP’s, get their vector from your character to that location, throw away the Z axis and apply that vector in the correct scale to your minimap.
Alternatively you could also just apply world coordinates to your minimap (so it’s aware which area it captures) and directly provide the coordinates to it.
There are some optimizations to not use a lot of resources since you’re doing this each tick.
For example you should just keep an array of “Location_BP” actors and update them as you get the begin and end overlap events. Also not every marker moves probably meaning you can provide this as well meaning the none moving once don’t have to be updated but only moved exactly like the minimap.