I followed Cameron's suggestion of snapping together the lateral weir and 2D flow area in GIS, then importing the GIS coordinates. Since I am modelling an entire reach in 1D-2D, my weirs are long (3-4km) which means that once I filtered the weir sta/elevation to 500 points, this created deviations between the weir elevation and the terrain - thus, I received errors that some 2D cells had higher elevation than the weir. Since this difference was small (~0.005m), I raised the weir elevation in Excel by 0.015/0.025 m, which resolved the error.

The model still won't run (1D/2D Flow Errors), but at least I don't have geometry errors (!)