My understanding is that, where a second Enhanced Filter uses output from a first Enhanced Filter, the first filter output is limited to 1000 issues.
As a user, I would like this limit removed, but if this is not possible, increased to at least 10,000.
This limit has caused me a lot of pain since our move from Jira Server to Jira Cloud. Whereas previously I had one saved Filter, I then had to artificially split this into around 20 parallel Filters, each referencing one of another set of 20 paralllel Filters. And I had to do this for around 5 saved Filters. So I now have 200 Filters to manage, which are now split across 4 users because of limits on Filter syncing (see https://productsupport.adaptavist.com/servicedesk/customer/portal/23/SRJCLSUP-4599).
All this has taken a considerable amount of my time and caused delays in our software release process. It worked just fine in Jira Server.
(By the way, I am not 100% clear on precisely where the 1000 limit applies - please do clarify/corrrect me!).