Or maybe he wouldn't have learned necessary and important lessons if he hadn't bought the textile business.
Sanborn Map company sold maps, but they also had an investment portfolio that dwarfed their market cap. So he had them sell it off. He did a lot of cigar butts like that. Back then, Buffett was pretty pushy. He would do whatever it took to acquire all the stock of companies and would sack the management teams and replace them with others. At one point an entire town in Dempster protested against him.
Then he met Charlie Munger and Munger got him to start focusing on acquiring good businesses that had franchise values. See's Candy is the best example. So he looked for companies with sustainable competitive advantages and high returns on invested capital. So he started to buy your good businesses run by great entrepreneurs with little actual control. The only thing he would ask is if they could not meet an internal hurdle for ROIC (return on invested capital) that they give excess capital for him to redeploy.
Now, if you look at what Buffett has been doing, he's continued to evolve a bit. He seems much more focused on buying utility-like businesses. MidAmerican and Burlington Northern exemplify this. They don't throw off cash that can be redeployed elsewhere like the previous Berkshire deals. At the same time though, they potentially can be around for the next 50 years and increase prices to make up for inflation.
These deals are probably going to continue. Buffett must realize that his 3 replacements wont be as good at allocating capital, so he will need to find businesses that are great but also utilize a lot of cash. That would reduce the burden on his CIO replacements.
PS: He once said, I can find plenty of people that can consistently turn 1 million$ into 2million$ in a year, I can't find people that can consistently turn 1 billion$ into 2 billion$ in a year.
He also has unprecedented access to financial data of companies he may be sniffing around - not the garbage you and I see in SEC documents. (although 10q documents are treasure chests of company info for research).
Goizueta did a lot of great things. He separated out the bottler business and got them to focus on returns on invested capital with every investment they made. He made a ton of money for shareholders. Buffett has remarked that he held KO stock too long but now I believe he keeps it mainly to collect the dividends, his cost basis is low.
Early years (during the hedge fund): 1) The "cigar butt" value-investor -- looking for companies which held more book value (assets like cash on hand, factories, etc.) than capitalization (stock price times shares outstanding). Arguably Sun Microsystems (right before Oracle bought them) and Yahoo now are pretty close to this in big tech companies, but back in early/mid 20th century, there were LOTS of these.
2) Value investing -- looking for brands which have growth potential but are undervalued by the market. Apple right after Steve Jobs returned would be a tech example.
3) Special case situations, like arbitrage -- for instance, right before a merger, buying one or the other stock, and if you have someone like Apple releasing a product which has a component from one vendor, buying that vendor's stock.
WB largely ended this strategy for two reasons: it was successful enough to use up most of the available opportunities (as well as attract copycats), and it was very labor/research intensive for the returns you could get -- good percentage returns, but it was difficult to employ large amounts of money. Sort of like angel investment vs. mezzanine IPO financing.
Mid years (70s and 80s): insurance
WB discovered the insurance industry (through GEICO), and basically had "free money" to invest (which is what an insurance company does with premiums). He went from value investing to buying entire companies (or large shares) which he thought would grow and remain good companies.
Late years: (1990s to now): top-quality He basically has the problem of having too much money to invest, and thus can only go after the largest opportunities; a 10x return on $1m is not enough to move the needle. Thus, WB and BH buy entire top-quality companies, with a focus on sustained performance, vs. making a lot of speculative bets.
The old-school WB (and even better, Ben Graham, his early role model), is the only interesting one to me -- once he had established his track record doing something well, he was in a position to make reasonable investments no matter what. Getting to that point was the interesting part.