This is both subjective and highly dependent upon the scope of services being run. My setup would probably progressively create more hassle than it saves as on a scale from small business to large business. For the setup I have, I quite specifically want to block their traffic.
I'm possibly overly militant about this, but they keep databases of the results of their scans, and their business is selling this information to ... whoever's buying. I don't want my IP addresses, open ports, services or any other details they're able to gather to be in these databases over which I have no control and didn't authorise.
To steal an oft-used analogy, they're taking snapshots of all the houses on all the streets and identifying the doors, windows, gates, and having a peek inside, and recording all the results in a database.
I believe all of them are illegitimate. They 'do' because they can, and it's profitable. "Making the internet safer" is not their raison d'être.
Happy for any else to form their own opinion, but this is my current stance.
"Nice network you've got there."
"We noticed something might be open. We're not telling you what it is."
"It would be a pity if something happened to your business."
"Give us lots of money."
Sounds like a movie strong-arm thug.
Act like a bot, get treated like a bot.
> Just be aware that with your strategy “blocking 50% of unwanted traffic” means blocking non-attack traffic
You don't block them forever, just enough for them to move on to someone else.
Note that you're basing your assertion on the motivation of random third parties exclusively on the fact that they exist and they are behind active searches for vulnerabilities.