Traditional evaluation studies of MANET routing protocols have concentrated on quantitative, traffic-based metrics like throughput and packet loss. These metrics provide a limited evaluation of protocol performance because they do not measure how well the protocol finds and maintains routes, only how well functional routes perform, once established. This work proposes a method for assessing how well a proactive MANET routing protocol tracks the network topology by comparing the reported routes to the actual topology. This study exposes the impact of routing message propagation and message loss on MANET routing by showing the existence of errant routes and protocol-reported information, specifically broken routes, incorrect routing table hop counts, and existing routes that were not found by the protocol. We use this approach to analyze the performance of the OLSR protocol in a medium-sized MANET, using data from the MANIAC Challenge. The results favor OLSR, but expose errant routes, how often they occur and for how long.