The recycling of waste plastics and agro-industrial byproducts for the creation of sustainable polymeric composites is acknowledged as an effective strategy to mitigate the harmful environmental impacts of plastic waste. Despite the considerable promise of sustainable composites in the Circular Economy (CE), their application remains minimal due to the absence of an appropriate material selection methodology. The presence of several contributing factors in material selection categorizes it as a multi-criteria group decision-making (MCGDM) issue. In contemporary MCGDM contexts, experts are frequently seen as independent agents, with their psychological behaviours rarely considered. However, in numerous cases, the viewpoints of experts are likely to influence each other, and these experts often demonstrate subjective bounded rationality in the decision-making process. The Pythagorean linguistic information (PLI)-based MCGDM model, utilizing the measurement, and ranking of alternatives via the compromise solution (MARCOS) approach and quantum decision theory, is designed to tackle this problem. The PLI consists of an ordered pair of a linguistic term (LT) and a Pythagorean fuzzy number (PFN), whereby the latter quantifies the degree of certainty regarding the former. PLIs, guided by sample size data, are utilized to aggregate group linguistic assessments, clearly demonstrating both quantitative variation and qualitative uncertainty. A thorough generalized Pythagorean linguistic MARCOS approach is presented to depict the constrained rational actions of decision-makers. It is employed to evaluate the superiority of each option. A quantum possibilistic aggregation framework is established to examine the interference effects among experts' perspectives. This technique considers expert opinions as concurrently existing wave functions that interact and influence the overall result. The comparisons with several contemporary methodologies confirm the effectiveness and reliability of the novel MCGDM strategy.