I think it's OK to make that mistake because "starting" is a more positive force than "starting over" is a negative one. if the abstraction or assumptions were

I think it's OK to make that mistake because "starting" is a more positive force than "starting over" is a negative one. if the abstraction or assumptions were wrong, you already built some kind of lossy approximation that is highly useful. its just important to recognize when the prototype needs to be thrown away or retired. the corollary of this, though, is that "there's nothing more permanent than a temporary fix". so, balancing these ideals is the work of engineering and management.