Is C # an interpreted language?

Many of the languages you listed are interpreted in the sense that, even though they are compiled to binaries, those binaries must go through a virtual machine that translates the code to machine specific code for the platform it is running on. This is a trade off of convenience and portability at the cost of speed.

C and C++ (like assembly language) are machine specific and are compiled down to machine code for the specific platform the compiler was told to compile for. Thus, they run native on that platform, thus no virtual machine translating overhead, thus much faster code. You have no portability (in terms of binaries) as you have to compile for each platform (which will likely require code rewriting for the specific platform) but, you gain a lot of speed.
You can find comparison of C++ (or C) to other languages on youtube regarding speed differences. And the speed difference can be massive. Such as running graphics on python (using raylib library) vs C++ (using raylib for C++ library). 14 frames per second vs 60 to hundred of frames per second difference for the exact same application using the same graphics library.
So, yeah, if there is a virtual machine involved I would still consider it interpreted as it is being translated in real/run time to local machine code, much the same as an interpreter (such as early versions of non-compiled Basic) is translating at run/time. This is overhead that a lot of apps can afford as they are not speed critical. But, for speed critical apps, this is not overhead that can be afforded.

Another difference is that a compiled language like C or C++ which produces native binaries use compilers that have very good optimizations. Not only can you optimize your code but the compilers will farther optimize and generally do a pretty good job at it. This is more easily done as the compiler only has to optimize for the platform your are specifically compiling for. Interpreted languages can only optimize to a certain degree as they have to produce binaries that need to be interpreted by virtual machines that follow specific standards... you can't optimize for specific platforms... you are producing generic code that has to run on a virtual machine interpreter.

As far as using C++ for web programming, yes it is used. It's not common unless you are in the field of making web servers or other such things though. But yes, it can be used;

Webserver in 40 lines of C++ code;

#include <nodepp/nodepp.h> #include <nodepp/http.h> #include <nodepp/path.h> #include <nodepp/fs.h> using namespace nodepp; int main( int argc, char** args ){ process::start( argc, args ); auto server = http::server(={ string_t dir = "www/index.html"; if( cli.path != "/" ) dir = path::join( "www", cli.path ); console::log( cli.path, cli.get_fd() ); if( !fs::exists_file(dir) ){ cli.write_header( 404, {{ { "content-type", "text/plain" } }} ); cli.write("Opps 404 error"); cli.close(); return; } auto str = fs::readable( dir ); cli.write_header( 200, {{ { "Content-Length", string::to_string(str.size()) }, { "Content-Type", path::mimetype(dir) } }}); stream::pipe( str, cli ); }); server.listen( "localhost", 8000, ={ console::log("server started at http://localhost:8000"); }); process::stop(); }