Wi-Fi is the preferred way of accessing the internet for many devices at home, but it is vulnerable to performance problems. In this work, we propose a method to estimate the link capacity of a Wi-Fi link using physical layer metrics passively sampled on commodity access points. We build a model that predicts the maximum UDP throughput a device can sustain, which extends previous models to consider IEEE 802.11n optimizations such as frame aggregation. We validate our model through controlled experiments in an anechoic chamber. Over 95% of the link capacity predictions present errors below 5% when using our model with reference data. We show how link capacity estimation enables Wi-Fi diagnosis in two case studies where we predict the available bandwidth under microwave interference and in an office environment.