This article presents an overview of existing surveys and reviews focusing on general and communication architectures (e.g., cloud- and P2P-based) within the metaverse. We propose GEIM, a green and economical multi-user interaction framework, to address inefficiencies in communication and computation, and to facilitate self-orchestrating and collaborative inter-actions. GEIM employs a flexible and intelligent communication and computation orchestration algorithm, along with caching strategy, to harmonize resources from remote cloud, edge cloud, and device layers for collaborative interactions. We deployed a GEIM prototype in real-world scenarios, demonstrating its capability to dynamically orchestrate interactive devices and computational resources, resulting in significant reductions in communication, computation, and memory overheads. Experimental results show that when GEIM collaborates with two edge nodes, it achieves a 17% reduction in GPU load, a 20% decrease in memory usage, and a minor 4% drop in frame rate compared to a single edge node setup. Last, we outline future research directions and conclude the research implications.