For your requirements with 3000+ tasks, complex sorting by process variables, and sub-second response times, the recommended architecture is the event-driven local database pattern rather than relying solely on Tasklist APIs. I found the following relevant resources:
- Understanding human task management - Deciding about your task list frontend
- User task listeners - Available in 8.8+ for capturing task lifecycle events
- Tasklist API versions - V1 deprecated in 8.8, V2 doesn’t support includeVariables yet
- Process variables and task availability issue - Community member’s production implementation using local database pattern
- Camunda 8 Tasklist UI - Slow UI performance - Performance considerations and hybrid approaches
- Data fetching - Variables - V2 API variable filtering capabilities
- Are there plans to add support for includeVariables in Orchestration Cluster API Search user tasks? - Current V2 API limitations
Does this help? If not, can anyone from the community jump in? 
Hints: Use the Ask AI feature in Camunda’s documentation to chat with AI and get fast help. Report bugs and features in Camuda’s GitHub issue tracker. Trust the process. 