The advent of the Networked Society has revolutionised the way public discourse is shaped and negotiated leading to increased access to both information and (dis)-(mis)information. With the goal of investigating the discursive roots of disinformation in crisis scenarios, we propose a scalable methodology to analyse polylogues where stakeholders (e.g. citizens, journalists, politicians) advance various positions (news claims) across multiple venues (e.g. social media, broadcast media, discussion fora). We apply this framework to the analysis of discourse(s) around the Events Research Programme (ERP) in the United Kingdom aimed at gathering evidence on the reopening of events and venues assessing the risk of SARS-CoV-2 transmission. Through the combination of qualitative (content analysis of survey results) and quantitative methods (topic modelling and sentiment analysis), we compare and contrast discourse(s) around the pilot live events from different stakeholders (UK government, official national and local media, social media users and aspiring participants at the large events) across media outlets (official website, news media outlets, Twitter and questionnaire). Drawing from attested mismatches we provide three main recommendations on how to craft effective public messages in crisis scenarios.