The thing about hacks is that they're going to happen. Yes, Amazon could be hacked so that your bank details would be exposed--but the bank can also be hacked directly. Up to now, hackers have done that mostly to steal money, but if they wanted info and couldn't get it in other places, they're more than capable.
The best practice at this point is to use an identity protection service. They do cost money, but they monitor and flag suspicious stuff for you. The good ones also provide funds (usually up to one million) to compensate for hacking losses, as well as support to help you restore your credit as quickly as possible.
Decent banks and credit card companies are also pretty good at flagging suspect transactions. In particular, if something unusually large (like a withdrawal in particular) comes through, they'll probably put a hold on it unless you're doing it in person in the bank.
For added protection, you can lock access to your credit records, as well as the ability to open a new utility account or bank account in your name. That doesn't protect you against a hacker trying to siphon money from existing accounts, but it does foil identity thieves trying to use your credentials to set up new accounts. I did that myself some time ago. Note that this isn't a good solution for everyone, particularly if you frequently need to take out loans or something like that. You can unfreeze things, but the transaction takes time.
A recent data breach (from a national database I've never even heard of) exposed millions of social security numbers, including mine. There's nothing to be done about that except to be vigilant. (Getting a new social security number is possible but creates a whole host of problems of its own.) Since I'm already locked up, I'm not too worried.
As far as Amazon is concerned, yes, it sometimes makes odd mistakes. But it also controls a huge chunk of the book market. In the US, 50% of all book sales (including all formats) take place on Amazon. For ebooks, which tends to be where the sales are for indies in most genres, the percentage is even higher, over 80% before Data Guy went behind a paywall. The picture is similar in the UK, Australia, and New Zealand. Even in Canada, where some people had the impression Kobo was bigger than Amazon, it was bigger than in the US, but a majority of ebooks were still sold on Amazon. These figures may have changed some in the last few years, but I doubt in a significant way.
I know people on Substack who are pursuing a non-Amazon strategy. As far as book sales are concerned, they appear to be doing worse than I am, though over the very long-term, that may change. Some of them do make subscription money on Substack. There's a large audience there, though considerably smaller than the purchasing population on Amazon. But building a subscriber base is a slow process unless one can import a huge fan base from elsewhere. I've been there since March 1 and have 221 subscribers, three of whom are paid. Two of those are old friends. My pattern is apparently not uncommon. Exceptions are people who give good writing and marketing advice. (Like everywhere else, subscriptions pour in for them. At least on Substack, the advice is actually good, from what I've seen. Some subscriptions also include coursework and even individual meetings with the advisor.) Also exceptions are celebrities, though on Substack, that means primarily famous journalists in a time when other journalistic opportunities are shrinking, political personalities, and trad pubbed authors. (I've interacted with at least four bestsellers, but there appear to be far more than that.)
Don't get me wrong--I like Substack, and it might provide supplemental income some day. But as a replacement for Amazon sales, it's unlikely. It doesn't hurt to try it out and see what happens. I've made some non-subscription money--but that was from getting Substack friends to buy books on Amazon. That actually works pretty well, sometimes better than AMS ads. But in general, subscription revenue seems to be a slow growth process for most of us.
