For better or worse, most commercial embedded development is done in C/C++. If you are already familiar with these, start with Atmel Studio and an AVR 8 bit processor (easier to learn the hardware side, simpler processor) or mbed.org and one of the supported boards (STM32 Nucleos are cheap and have built-in JTAG programmer/debugger with a wide range of processors).
Pay could be higher or lower, you will almost certainly take a cut initially but there are plenty of high paying embedded development jobs.
We have been using Asterisk (SIP/IAX) for about 7 years both on "cloud" servers and field communication devices with no issues on production devices (only minor problems during development/qualification). Our experience has been very positive and the community is very helpful.
If you are looking for speeds on the order of 0.0024 Mbps (2400 baud) then maybe. You could use amateur radio packet radio gateways and radio modems and set something up.
You could do it legally (in the US) by paying a token amount and getting a part 90 business license or possibly by using MURS frequencies instead of FRS/GMRS.
>paying a token amount and getting a part 90 business license
The spectrum is reserved for actual business use; "amateur radio, but with encryption" is not on the list of eligible use cases for a business band license [0]. If you're willing to break the law, wouldn't it be simpler to go ahead and transmit encrypted on the amateur spectrum than to submit a fraudulent application for a business band license?
Take a look at Autodesk Fusion 360. It is free to makers / small startups. I use it extensively for product prototype development including for 3d printing and subtractive machining. Another good resource is McMaster Carr, they have downloadable 3d models for most of what they sell. Having accurate 3d models for all the fasteners and other parts makes a big difference in quickly building a 3d model as well as the quality of renderings done in the cad system.
Get a good lawyer, even if you only pay for a couple of hours. There are things you can do immediately (depending on whether you received a letter or they actually filed a suit) that may make them go away.
I was party to one of these a couple of years ago and if we had received good legal advice early it would have been much less expensive in the long run.
As a developer it is worth it as long as you have separate teams responsible for development and production. Segmentation means that a prod server can't accidentally be configured to access a dev resource (db instance, etc). If you have development hand off installation packages and instructions it also means that dev never has to touch prod servers (except read access to logs, etc).
It also eliminates the temptation to "temporarily" use a dev machine/instance as part of prod to solve a problem (and then leaving it that way for years).
I would start with Octave. You can use it to quickly learn the math behind music synthesis and experiment with the various methods. It is open source and available for Windows,Linux,Mac and Android (at least, there are probably other ports).
Then you can concurrently learn a programming language (IMHO based on your post, Python might be a good place to start).
You might also take a look at CSound which is a dedicated environment and programming language for sound synthesis. The CSound documentation pages also have references to a wide variety of other available systems (like Max and Reaktor).
If you are going to ultimately create audio apps for the iphone then I don't think you can avoid C/C++/Objective C but you will have an easier time doing the sound synthesis experimentation in an environment tailored to that function.
