You've taken an approach here where you intentionally hide the fact that a video file is (at least) 3 different technical formats that are independently variable. An "MP4" file can have a range of audio codecs, a range of video codecs, timed-text formats, additional audio, and so on. And there is no single standard composition that works everywhere.

When you simplify a matrix of user choices by making the vcodec, acodec, and timed text format choices on behalf of your users, you take on the burden of making sure those work everywhere the users want to playback. What you'll find is that most devices on the market only support a very limited range of container+vcodec+acodec combinations, they are undocumented more often than not, and buggy as hell.

The oversimplification approach you're taking is "ingesting anything, but output only 'Value Meal #1' for everybody." This has value for some people, but it puts a big burden on you to make choices that playback mostly correctly on a wide variety of devices, and it mysteriously breaks things that don't work everywhere (like surround sound, ambisonics, many timed text formats etc.). There's a reason why all that choice exists, even if most people don't, don't want to, and shouldn't need to understand it.

Not trying to dissuage you. Just sharing experience. :-)