Much like MPEG-21, Dolby AC-4 is modern multi-media format designed to accommodate the needs of media broadcasters. However, unlike MPEG-21 - which is almost completely unknown by consumers - Dolby AC-4 is known to some households, and is likely to become more well-known over time. If there's one thing besides psychoacoustics that Dolby is really good at, it's marketing. Furthermore, AC-4 is a codec, while MPEG-21 is a container. They are both of interest to the broadcasting industry, but for different reasons.
Dolby AC-4 is simply the officially recommended audio component of another multi-media standard most people have never heard of: ATSC 3.0. ATSC stands for the Advanced Television Systems Committee, a broadcast television standards committee consisting of industry players established in the early 1980's to establish standards around digital television. Ironically, although a U.S. based organization, its guidelines were adopted much more rapidly by foreign corporations. The U.S.A. did not truly start catching up in the digital TV space until the early 2000's.
ATSC 3.0 is simply the 3rd generation of ATSC digital broadcast standards. The digital broadcast industry moves very slowly relative to many other industries. Changes require substantial capital investment by its players (e.g. TV stations and other large digital broadcasters such as DirecTV). In the United States, ATSC 3.0 incorporation is still rare. This is largely due to the need for specialized hardware on the receiving end of content distribution (i.e. for consumers). Nearly all current TVs in U.S. homes will require some sort of decoder box to use it. So, while the FCC approved guidelines for ATSC 3.0 usage in 2017, its adoption in the United States has been muted so far. Regardless, this will change at some point.
From an auditory perspective, ATSC 3.0 supports two (2) multi-channel audio formats: MPEG-H and Dolby AC-4. Since 2017, AC-4 has dominated most broadcaster plans in terms of support. I suspect this is largely due to Dolby's household name recognition rather than it being superior to MPEG-H.
So, what's the big deal with AC-4? Well, first off consider the fact its predecessor (at least in name) - Dolby AC-3 - is nearly 30 years old (circa 1991). Of course, Dolby has released a dozen newer standards in surround sound/multi-channel sound since AC-3, but none of them were designed specifically for broadcast usage. Ironically, as old as it is, AC-3 works very well in up to 5.1 channel sound systems, and has proven to be a surprisingly versatile solution for virtually all needs (including broadcast multi-channel). One may think of AC-3 like MP3 files. Yes, the standard is really old, but it's still dominant because it's just that damn good. If it ain't broke, why fix it?
What is happening in modern times is psychoacoustics research and technological solutions have made tremendous progress in solving some very tricky use case scenarios around both recording and faithfully reproducing surrealistic multi-channel surround sound. Just think about Atmos for example, which debuted in 2014. A lot has changed since 1991, and much of that change has come in the last 5 years or so. Welcome to the world of 3D Audio. Specifically, three-dimensional audio where you don't need a specially designed room to experience the 3D effect (though it's still going to feel more realistic if you do). Dolby AC-4 and MPEG-H are both designed especially for recreating immersive audio experiences.
Another piece of the puzzle is while amazing audio capabilities have been defined during this time, more data means more bandwidth required to deliver this information digitally. And while it's true this is also less of a concern than 30 years ago, ironically there are still numerous valid concerns over low network bandwidth restricted devices' (e.g. cellular phones) ability to consistently receive digital data in such a way to facilitate the desired end-user experience. Given this world of conflicting priorities we find ourselves in, this places a lot of requirements on any new multimedia standard. Not only must it look and sound great, but it also needs to be stream-able across a wide variety of hardware platforms, some of which are notably network bandwidth-limited.
As such, AC-4 is intended to combine improved audio quality in conjunction with reduced data bitrate requirements. You may think of Dolby AC-4 as a project designed to apply a Dolby Digital Plus model experience to users on multiple types of platforms. One of the key components of AC-4 is its ability to negotiate the volume and depth of data between the sending and receiving devices. This makes AC-4 ideal for platforms such as mobile phone carriers, where the capabilities of the end user's device to reproduce high quality audio and the end-point device's potential network bandwidth are both likely limited. Previous codecs presume both source and playback devices are local and data throughput is not a challenge. They also place the onus of downgrading or tossing extraneous content upon the playback device. This means in traditional audio playback environments, the playback device must be capable of accepting audio data at least as rapidly as it is doled out by the source device. AC-4 changes this dynamic, by making the data transfer a negotiation process that takes the playback device into consideration and makes the playback device's capabilities a priority. This means down-conversion is processed by the sending device. Overall, AC-4's more efficient architecture promotes the ability to play any content on any device, anywhere.