An individual can make a consistent argument with properly obtained data and conduct analysis according to standard methods and make a very wrong conclusion.
Robustness is achieved by reproduction of a particular result despite analysis based on heterogeneous theory, source data, and method. This is most feasible at the level of a social collective rather than an individual. A well meaning scholar can at beast achieve an inferior level of robustness.
An individual may make their own argument more robust by employing alternate types of analysis, accruing data from multiple sources, and so on, but this can rapidly become temporally or financial infeasible. It is efficient to require each scholar to make their own argument maximally robust because the efficient case is where marginal cost equals marginal benefit, but marginal cost generally exceeds marginal benefit when the scholar is forced to maximize robustness.
The well meaning scholar can improve the robustness of their argument to a point, but they will run into certain constraints which cannot be overcome even with additional time and money. Specifically, there is a risk that they will unconsciously employ consistent basic assumptions or cognitive biases which will systematically alter their results, perhaps even predetermining a false result.
The first scholar might hire another scholar to conduct independent analysis and hopefully eliminate some of these biases, but the hired scholar will not truly be independent because there will be an incentive for them to tell the first scholar what they want to hear. This arrangement is also expected to be less efficient than if there was a more flat organizational structure in which the first and second scholars were competitors.