Being correct on an issue is often tempered by the necessary clown-car and selection filter of consensus agreement. It is important to find rational agreement and indeed science could not be credible without competent peer-assessment but the problems of democratic consensus include that:
1) only those assertions considered rational from within a current paradigm, semantic ontology and predominant conceptual vocabulary are considered worth consideration or investment – thus already and implicitly privileging those (bureaucrats, lawyers, administrators, etc.) who have mastered the contextual lexicon du jour; and,
2) the effect of statistical aggregation tends to even out accepted results into a simplest, most easily agreed-upon solution and clustered sensibility towards the center of the bell curve (of aggregated opinion as much as of fact) – outlying results and revelations are unnecessarily devalued and administratively obfuscated by virtue of their eccentric location on the curve.
Being correct is just never enough when everyone else also thinks they are correct, within a widely divergent margin of error. In an era when almost everyone has access to publish or broadcast an opinion, the extent of accuracy or fidelity to truth and fact across distributed information cultures are subject to wild variations.