Yes, I think Ghidra is "good enough" for most use cases. For some (architectures other than x86/ARM), it is vastly superior. For others (Windows and especially

Yes, I think Ghidra is "good enough" for most use cases. For some (architectures other than x86/ARM), it is vastly superior. For others (Windows and especially C++) it's not quite as good, but still useable. For C++ projects targeting Windows, IDA is still a lot better in terms of decompilation quality. The demangler and function signature system is particularly tailored to MSVC++. For other architectures, Ghidra kills IDA by virtue of abstraction - all disassembly turns into the same IR (P-code) and is decompiled from there. So, the decompiler works across all architectures with a processor spec, rather than being specialized to x86/ARM. And, the exposure of this P-code system is very powerful as well - with some clever plugin scripting for Ghidra, you can actually single-step through the P-code as though you had a full emulator/debugger, without needing to set up an entire virtual environment. https://wrongbaud.github.io/posts/kong-vs-ghidra/ https://wrongbaud.github.io/posts/kong-vs-ghidra/