Serverless computing has emerged as a prominent cloud application development paradigm over the past decade, enabling developers to deploy and execute applications without managing underlying infrastructure. By abstracting resource provisioning, configuration, and auto-scaling, serverless platforms leverage Backend as a Service (BaaS) and Function as a Service (FaaS) models to execute applications as independent, event-driven functions on managed infrastructure. In this paper, a set of well-defined research questions are formulated to systematically analyze the challenges, design trade-offs, and performance implications of serverless computing. A comparative evaluation with traditional cloud models, including Infrastructure as a Service (IaaS) and Platform as a Service (PaaS), is conducted with respect to cost efficiency and response latency. Beyond existing surveys, this work provides a unified and quantitative synthesis of serverless research by jointly examining architectural features, execution workflows, performance metrics (e.g., cold-start overhead, throughput, and cost-effectiveness), and the availability of frameworks and simulators within a single analytical framework. Unlike prior studies that focus on isolated aspects of serverless systems, this review consolidates insights across multiple dimensions, bridges application and system level perspectives, and identifies underexplored research gaps such as stateful serverless execution, benchmarking reproducibility, and simulator-driven evaluation. The paper concludes by outlining emerging research opportunities, offering a comprehensive and up-to-date reference for researchers and practitioners in cloud and serverless computing.