Greedy-only decoding is even worse. There's a reason every public model comes with suggested sampling parameters. When you don't use them, output tends to degrade severely. In your case simply running a 14B model on the same hardware with the tools you compare against would probably be both faster and produce output of higher quality.
Speculative decoding is the same as speculative execution on CPUs. As long as you walk back on an incorrect prediction (i.e. the speculated tokens weren't accepted) then everything is mathematically exactly the same. It just uses more parallelism (specificslly higher arithmetic intensity).
This should maintain exactly the quality of the original model, no?
Greedy decoding means it is still not ready though.
I've never heard of ANY speculative decoding that wasn't lossless. If it was lossy it'd be called something else.
This page is just a port of DFLASH to gguf format, it only implements greedy decoding like you said so the outputs will be inferior, but not inferior to greedy decoding on the original model. Tho that's just a matter of implementing temperature, top_k, etc.