This paper aims at investigating further the use of the social media Twitter as a real-time estimator of the US Air Transportation system. Two different machine learning regressors have been trained on this 2017 passenger-centric dataset and tested on the first two months of 2018 for the estimation of air traffic delays at departure and arrival at 34 different US airports. Using three different levels of content-related features created from the flow of social media posts led to the extraction of useful information about the current state of the air traffic system. The resulting methods yield higher estimation performances than traditional state-of-the-art and off-the-shelf time-series forecasting techniques performed on flight-centric data for more than 28 airports. Moreover the features extracted can also be used to start a passenger-centric analysis of the Air Transportation system. This paper is the continuation of previous works focusing on estimating air traffic delays leveraging a real-time publicly available passenger-centered data source. The results of this study suggest a method to use passenger-centric data-sources as an estimator of the current state of the different actors of the air transportation system in real-time.