Airtable is not intended (or designed) to be your data back-end. Some people use it as such, but it’s hardly the likes of services such as Firebase, so it has strict limits. And even if they allowed 10 requests per second, it’s still running inside the container that represents your Airtable instance. If anything, this limit is to defend you from your own users; it is not some arbitrary penalty threshold designed to protect Airtable services or other Airtable customer instances.
My recommendation - consider caching forward your data using something like SynInc to provide a near limitless and performant architecture for Airtable data access.
I’ve built a product that could potentially help you out here. It’s called SheetDream (https://sheetdream.io). Among other things, it can essentially act as a buffer in front of Airtable (using a caching DB) to help alleviate rate limit issues.
Get in touch with me, we can see if our service might be a good fit for you. Email: firstname.lastname@example.org