End-of-Life (EOL) products in remanufacturing are prone to a high degree of uncertainty in terms of product quantity and quality. Therefore, the industrial shift towards a circular economy emphasizes the need for agile and hybrid disassembly systems. These systems feature a dynamic material flow. Besides that, they combine the endurance of robots with the dexterity of human operators for an effective and economically reasonable EOL-product treatment. Moreover, being reconfigurable, agile disassembly systems allow an alignment of their functional and quantitative capacity to volatile production programs. However, changes in both the system configuration and the production program to be processed call for adaptive approaches to production control. This paper proposes a multi-priority rule heuristic combined with an optimization tool for adaptive re-parameterization. First, domain-specific priority rules are introduced and incorporated into a weighted priority function for disassembly task allocation. Besides that, a novel metaheuristic parameter optimizer is devised to facilitate the adaption of weights in response to evolving requirements in a reasonable timeframe. Different metaheuristics such as simulated annealing or particle swarm optimization are incorporated as black-box optimizers. Subsequently, the performance of these metaheuristics is meticulously evaluated across six distinct test cases, employing discrete event simulation for evaluation, with a primary focus on measuring both speed and solution quality. To gauge the efficacy of the approach, a robust set of weights is employed as a benchmark. Encouragingly, the results of the experimentation reveal that the metaheuristics exhibit a notable proficiency in rapidly identifying high-quality solutions. The results are promising in that the metaheuristics can quickly find reasonable solutions, thus illustrating the compelling potential in enhancing the efficiency of agile disassembly systems.