It's probably far, far worse than that. How do you count things like when someone suggests using a database/structure/system that is very poorly suited and spen

It's probably far, far worse than that. How do you count things like when someone suggests using a database/structure/system that is very poorly suited and spends 3 months or more before giving up versus someone that can figure out ahead of time it won't work? Or when someone suggests an approach that is error prone and continually introduces faults, yet had a slightly lower startup time? Or when a system is designed and needs 20x the hardware? Or requires literally 1000x the processing? While in some very constrained, straightforward coding challenges it might be hard to find orders of magnitude improvement, a lot of designing software has a huge range of solutions. It might be the case that a single org doesn't get to compare.