In Radio Frequency (RF)-powered networks, peak antenna gains and path-loss models are often used to predict the power that can be received by a rectenna. However, this leads to significant over-estimation of the harvested power when using rectennas in a dynamic setting. This work proposes more realistic parameters for evaluating RF-powered Body Area Networks (BANs), and utilizes them to analyze and compare the performance of an RF-powered BAN based on 915 MHz and 2.4 GHz rectennas. Two fully-textile antennas: a 915 MHz monopole and a 2.4 GHz patch, are designed and fabricated for numerical radiation pattern analysis and practical forward transmission measurements. The antennas' radiation properties are used to calculate the power delivered to a wireless-powered BAN formed of four antennas at different body positions. The mean angular gain is proposed as a more insightful metric for evaluating RFEH networks with unknown transmitter-receiver alignment. It is concluded that, when considering the mean gain, an RF-powered BAN using an omnidirectional 915 MHz antenna outperforms a 2.4 GHz BAN with higher-gain antenna, despite lack of shielding, by 15.4× higher DC power. Furthermore, a transmitter located above the user can result in 1× and 9× higher DC power at 915 MHz and 2.4 GHz, respectively, compared to a horizontal transmitter. Finally, it is suggested that the mean and angular gain should be considered instead of the peak gain. This accounts for the antennas' angular misalignment resulting from the receiver's mobility, which can vary the received power by an order of magnitude.