Pixel density vs megapixels for distant architectural detail

Shot a temple roofline from across a river last month and the crop didn't hold up the way I expected from a 45MP file. Got me thinking about whether a high density crop sensor body would actually resolve more of that distant detail than my full frame does.

For anyone shooting architecture, street scenes, or anything where you can't always get physically closer, have you compared a crop sensor with tight pixel pitch against a higher megapixel full frame? I'm mostly after fine texture in stone, tile work, signage. Not wildlife distances, but far enough that I'm leaning hard on the crop.

Curious what bodies people are actually using for this and whether the real world difference matched what the specs promised.