Sam, you're repeatedly making an absolutely huge, gigantic mistake. You are making this mistake every time you say "most", "normally", "usually". Huge.

Because the world of ambitious, information technology (IT) start ups is necessarily looking for something exceptional which, necessarily has next to nothing to do with "most", "normally", or "usually". What happened in the past, that is empirical probabilities, even for the successful start ups, is next to irrelevant for what will be powerful, valuable, and new tomorrow. Did I mention "new"? New can very much mean different and not part of most, what is normal, or what is usual.

Or, take a famous seccess story, e.g., Cisco, Intel, Microsoft, Sun, Oracle, Google, Facebook, etc., review the environment when they were getting started, and, then, find for each of those successful companies how much they looked like most, normal, or usual. Answer: Not much, and where they were "most", etc., it was essentially irrelevant.

Or, the SR-71 was nothing like most, normal, or usual. So, how to evaluate the project? Sure: The evaluation was no doubt quite technical and thorough but also quite accurate. No doubt in this evaluation most, normal, and usual played no role.

Sam, looking at most, normal, and usual in the past will tell you next to nothing about what will be successful in the future; still for a wide class of start ups, there are ways to tell.

Sam, more simply, clearly you are looking at most, normal, and usual as guidance for what should look for or do for the future, but, especially for the criteria you are considering, this is poor and will throw out a larger fraction of good babies than bath water.