On recent C++ projects, I used the TAP [0] test format. It's quite minimalistic and does the job.
Debugging tools have improved quite a bit. I suggest you take a look at asan [1] for memory related errors. Also, tsan [2] looks very promising for data races detection. I think both tools have been integrated within LLVM/Clang/GCC.
Also, you might want to take a look at tracing tools such as (shameless plug!) LTTng [3] for userspace.
Static analysis tools are quite "en vogue" right now. IMO, nothing comes close to Coverity Scan [4]. It's free for open source projects, but you will have to fork (quite) a bit of money to run this tool on proprietary applications. On the other hand, purely FOSS static-analyzer such as clang-analyzer and cppcheck can catch nasty errors. I would suggest using a mix of all these tools.
I respectfully disagree with this. Ignoring return values is _not_ good practice. It is a slippery slope to bad software. By catching these memory errors, a program has the chance to properly teardown and report a message to the user instead of crashing.
Absolutely! Codebases in "franglais" are terrible. Also, when documenting such codebases, I cringe whenever I see public APIs with mixed language members.
[1] - https://github.com/brendangregg/FlameGraph