I'm generally speaking of the stage right after a startup closes an institutional seed round. Across 75+ such investments over the past 10+ years, I've seen gaps in each of those areas be catastrophic. It definitely doesn't require "a few dozen employees" to mess up.
Especially on the operations side, I think engineers underestimate the amount of stuff there is to do. Have a distributed team? Welcome to state taxation hell. Have some pilots? You're going to have to deal with a lot of NDAs, cooperation agreements, onerous contracting requirements, building trust by connecting advisors (but not using them too much), etc. Mess up your banking and miss payroll? Your best employees will leave. Fail at reference checks during hiring and hire someone who results in a sexual harassment allegation? Your investors will have your head.
I think you're also misunderstanding what I mean by the product skillset. There are plenty of founders out there who can write amazing backend code, but couldn't deliver a usable product to their customers if their life depended on it. Sometimes that's ok, but most of the time being able to deliver a useful product can work even if your backend engineering is lacking -- not so the other way around. Ideally your technical co-founder can do both, but sometimes, especially when all the founders are engineers, a product-oriented CEO and an engineering-oriented CTO can work really well.
While sales is important, it really should be all the founders learning it, and good investors will go through great pains to help, either by sending founders to "sales school", recommending specialized advisors, or even pitching in themselves, using their connections to get you in the door at "big logos". In my experience, having a founder whose primary skill is sales can either be a positive or negative, depending on what else they bring to the table. If they bring nothing else, it's usually a negative.
At the stage of an institutional seed round, I'd generally agree with you. At that point you're big enough that screwing up your customer experience will hurt you.
It's the early-founder stage, where you're still proving out an idea and verifying that people will use it and likely will have to pivot several times, that I think sales + engineering (or ops if technology is not your core competency) is key. The advantage of having a really good engineer over a product person is that a really good engineer can "hit the high notes" and deliver a product that most ordinary people assume is impossible.
Especially on the operations side, I think engineers underestimate the amount of stuff there is to do. Have a distributed team? Welcome to state taxation hell. Have some pilots? You're going to have to deal with a lot of NDAs, cooperation agreements, onerous contracting requirements, building trust by connecting advisors (but not using them too much), etc. Mess up your banking and miss payroll? Your best employees will leave. Fail at reference checks during hiring and hire someone who results in a sexual harassment allegation? Your investors will have your head.
I think you're also misunderstanding what I mean by the product skillset. There are plenty of founders out there who can write amazing backend code, but couldn't deliver a usable product to their customers if their life depended on it. Sometimes that's ok, but most of the time being able to deliver a useful product can work even if your backend engineering is lacking -- not so the other way around. Ideally your technical co-founder can do both, but sometimes, especially when all the founders are engineers, a product-oriented CEO and an engineering-oriented CTO can work really well.
While sales is important, it really should be all the founders learning it, and good investors will go through great pains to help, either by sending founders to "sales school", recommending specialized advisors, or even pitching in themselves, using their connections to get you in the door at "big logos". In my experience, having a founder whose primary skill is sales can either be a positive or negative, depending on what else they bring to the table. If they bring nothing else, it's usually a negative.