Deploying fifth-generation (5G) networks in emerging markets demands a balance between performance targets and constraints in budget, spectrum, and infrastructure. We use MATLAB simulations to quantify how radio and architectural levers - MIMO (beamforming, diversity, spatial multiplexing), carrier aggregation (CA), targeted spectrum refarming to New Radio (NR), mmWave propagation with blockage/rain, and Non-Standalone (NSA) versus Standalone (SA) cores - affect capacity, coverage, latency, and interference robustness, with D2D and M2M as complements to wide-area access. Beamforming improves cell-edge SNR by about 3-6 dB, while spatial multiplexing dominates at moderate/high SNR via multi-stream gains. Throughput scales strongly with CA: increasing from 1 to 5x20-MHz carriers raises peak rate from about 200 Mb/s to about 1 Gb/s at 30 dB SNR; water-filling adds 5-12% over equal power at mid-SNR. Targeted mid-band refarming to NR increases median throughput by 60-90% in urban and 40-70% in rural scenarios when sub-1-GHz layers preserve coverage. At 28 GHz, rain and human blockage add about 8-30 dB excess loss, so viable mmWave deployment concentrates in LOS hot zones with narrow-beam arrays and short inter-site distances. NSA delivers broader initial coverage than SA by reusing LTE/EPC, while SA becomes attractive as transport improves (e.g., >= 10 Gb/s and < 5 ms RTT) and site density grows. We synthesize these results into a practical roadmap: start NR on NSA, prioritize CA-centric spectrum strategies with focused refarming, densify selectively in demand hotspots, and migrate to SA as backhaul and device ecosystems mature.