No, the reason is that it required substantial financial investments, and in some cases access to proprietary big-data collections.
mnist might have required a large amount of data at its creation, but it became a staple dataset
There was a lot of evolution before ChatGPT