What's the loading like? If it's an angle the loads are likely tiny and I suspect the plate is fine in compression in almost pure bearing through the plate and into the concrete without much bending actually happening in the plate. If you want to assume equal bearing under the plate, fine. Find the furthest perpendicular distance between a plate edge and a leg of the angle and use that for your plate cantilever distance. Use the full plate width for resistance width. Conservative, but a quick calc and your plate is likely still going to be thin given the loads.
For tension measure from the bolt to a reasonable point on the angle (will depend on the orientation) and make an assessment based on a conservative width in bending.
For low load applications where the arrangements are awkward to do the math on, just make conservative assumptions that keep the math simple. You're not going to save money on the steel compared to the amount of work you put in.
Realistically, the angle is likely going to force a hinge in the plate at the diagonal between the two tips of the angle. You also can't fail a plate without the hinge going to the edge at some point. So bending width equal to the plate width is likely conservative for a reasonably size, reasonable thickness plate.