Yes, such a replacement is inevitable - eventually. But it's important to understand (a) what 4G is, and (b) the dynamics of the evolution of wireless technologies as applied by the carriers. To begin, a "G" is a somewhat imprecise term, although there are sometimes formal definitions from recognized bodies like the International Telecommunications Union (ITU). 1G, for example, was analog (there are no more analog services, as far as I know, operating anywhere in the world today), 2G is digital but narrowband (GSM is the classic example here), 2.5G added packet data (known as GPRS) to 2G, and 3G is digital, packet, and broadband for data, but still circuit-switched (TDM) for voice. The formal definition of 3G, from the ITU, extends from 144 kbps to 2 Mbps, and the technologies include EDGE, 1XRTT, UMTS, and Ev-DO.
We recently received from the ITU the formal definition of 4G, which covers the range between 100 Mbps and 1 Gbps. Now, no one will actually get throughput anything like that, of course, but of even greater interest in the 98-Mbps gap between 3G and 4G. Many analysts refer to services in this zone as 3.5G, 3.75G, and even 3.9G, but the marketers at the carriers generally call these 4G. These technologies include HSPA, HSPA+, WiMAX, and LTE. It's possible, though unlikely that one of the other technologies developed in this zone, and there are a few, may emerge commercially as well.
It should be clear from the above that newer, more capable and cost-effective technologies always replace their predecessors, but such always takes many years. 3G took about seven years from commercial introduction to reach critical mass, which I define as a level of availability, service, and performance that a business user can depend upon, and at a price business users can justify. 4G is in its early days; all 4G products fall back to 3G given today's lack of the geographic coverage that is essential to critical mass. Prices of 4G can, similarly, be quite high. But that's in fact the key to making the decision - it's all about the ROI. If one can demonstrate a sufficient economic return on investment over the useful life of a given purchase, and the opportunity cost (the price paid vs. the cost of the next-best option) is acceptable, then go for it. Yes, we will all be using 4G eventually. But I suspect that, for now, most users will stick with 3G while happily getting both their jobs done and sufficient ROI.