This paper investigates the statistical-queueing constraints and pilot contamination phenomenon in irregular cellular massive multiple-input-multiple-output (MIMO) systems where base stations are Poisson distributed. Specifically, analytical expressions for the asymptotic signal-to-interference-ratio (SIR) coverage, rate-coverage and effective capacity under the quality of service (QoS) statistical-exponent constraint are provided for uplink transmission when each base station deploys a large number of antennas. We show that the QoS constrained capacity is in proportional to the path-loss exponent and inversely proportional to the pilot reusing probability which in turn is a function of cell load. Simulation results prove that pilot reuse impairments can be alleviated by employing a cellular frequency-reuse scheme, e.g., with frequency reuse factor (FRF) of unity, we see that, 40% of the total users have SIR above -10.5dB, whereas, with frequency reuse factor of 7, the same fraction of users has SIR above 20.5dB.