For point (6) about the capital, an argument can be made that large amounts of capital aren't necessarily bad. They can also lead to massive moats. Many great companies boil down to being able to significantly reduce costs over time. For example, Carnegie and Frick invested heavily in capital expenditures so that they could reduce costs over time. Once this happened, they could outprice the competitors leading to more money for Carnegie and Frick which lead to them being able to reduce costs further. A similar example nowadays is TSMC. They spend 36B on CAPEX in 2022.
Otherwise, if your business relies on scraping other peoples data and then presenting a nice UI around that, it could just as easily be replicated by others (no moat, as we're saying nowadays).
Of course, much like the niche chemical salesman, maybe the ideal scraping business would be to find such a niche that is small enough that other people wouldn't find it worthwhile to replicate your SaaS.
And even behind a login can still be scraped. (The proverbial "If I can do it anyone can" applies here along with "where there is a will there's a way")
Really excellent read. Very amusing to see his analysis of Apple and Microsoft, which seemed basically accurate until the iPhone came out.