One of the main challenges that the upcoming sixthgeneration (6G) wireless networks will encounter is the necessity to design sophisticated multiple access techniques that besides being capable of supporting massive connectivity, they can also fulfill the heterogeneous requirements of 6G services, namely further-enhanced mobile broadband (feMBB), extremely reliable and low-latency communication, and ultra-massive machine-type communication (umMTC). To this end, this work investigates the coexistence of multiple feMBB and umMTC wireless sources in a network. In order to enhance the achievable connectivity, each orthogonal resource block of the network is assigned to one feMBB and multiple umMTC sources. FeMBB sources are assumed to constantly transmit, while umMTC are considered to access the network in a probabilistic manner. If more than one umMTC sources attempt to access the network in a given resource block, no umMTC transmission is permitted, however, when precisely one umMTC source endeavors to access the medium, rate-splitting multiple access is employed to concurrently serve both feMBB and umMTC transmissions. For such a communication scenario, we derive novel closed-form expressions for sources' outage probabilities (OPs), ergodic rates (ERs), system throughput, and ergodic sum rate under both the ideal case of perfect channel state information (pCSI) and perfect successive interference cancellation (pSIC) and the more realistic scenario of imperfect CSI (ipCSI) and imperfect SIC (ipSIC). Furthermore, a high signal-to-noise ratio analysis is provided revealing deeper insights for sources' asymptotic behavior under all considered cases. Simulation results corroborate the accuracy of the derived analytical expressions, investigate the impact of different system parameters on sources' OP and ER performance, and illustrate the detrimental impact of ipCSI and ipSIC on system performance compared to the ideal case of pCSI and pSIC.