Only you can decide whether or not it's worth it. There are a few situations that you should consider:
How does your typical user visits your site? Are they primarily visiting because you gave them physical business cards or adverts where they have to type your domain manually? Then you should register and redirect all common typo domains to your main domain. Are your users primarily visiting you from online adverts or search engine? Then you probably don't need to worry much about typo domains.
Are you aware of overseas businesses in similar industry that have similar sounding names that might one day expand to your country? Then you might want to consider purchasing their domain name to protect your own trademark.
Are you using a country specific TLD? Are you not on the .com.xx of your country's TLD? Are your main domain on a novelty TLD? Then you might want to consider purchasing the .com version of that domain name, in case you have users that thinks that every website ends with .com, or if they tried to use Ctrl+Enter on the URL bar.
Are there other countries that your business might plan to expand to in the near future? Then buying your name under that country's TLD might be something you might want to consider.
Does your main domain names consist of common, regular words in the language in your primary audience? Is there common, similar sounding/similarly typed words or synonyms that your audience might misremember your name as? Does your name consist of unusual words?
One thing you definitely want to do to set a budget limit on your domain purchases. The number of you would have to have for this kind of typosquatting are endless. There are endless variety of permutations and you cannot purchase every single permutation of typos and similar names unless you have endless budget, which I assume you don't. So you'll need to set priority of what matters to you most. Domain names aren't that expensive, but they do cost quite a bit, and you need to renew them every year or so. If you're a smaller business, holding 500 typo domains are going to eat into a good chunk of your revenue. If you are Google, these expenses are negligible.
How much are you willing to spend to hold an asset that don't actually make you any money, for a situation that probably won't matter much even if it happens. Do you have other ways you could better spend the money? That $100/year you spend on squatting might be better spent on upgrading to a higher tier server or to buy a CDN service to help improve that performance issue that you've been having.
buysteal.comto that list?