The paper conducts an evaluation of traffic volumes, delivery ratios, and delays under various demand response (DR) setups using two distinct wireless solutions for smart grid (SG) communications. The first solution considers public long term evolution (LTE) network and the second one considers cluster-based hybrid sensor-LTE network where wireless sensor network (WSN) clusterheads are also equipped with LTE remote terminal units. The DR scenarios reflect cases where certain percentages of end users take part in automated DR-based load balancing while the rest of the users resort to advanced metering infrastructure based energy monitoring. In the DR cases various setups of energy consumption monitoring and load-balancing feedback are simulated. The purpose is to identify the limits of reporting and feedback intervals of the two communications technology setups for DR operations. DR and its management are becoming more important as distributed energy generation becomes more popular in households due to reducing prices of small-scale renewable energy generation equipment. The use of public telecommunications infrastructure is a good candidate for enabling DR communications over SGs, but it is becoming more congested by the increasing mobile data usage of consumers. The results show that both of the solutions have their advantages, LTE communications generally providing a higher delivery ratio whereas, surprisingly, hybrid sensor-LTE communications generally provides lower uplink delay.