I think it is worthwhile to explore cryptography that is not based on the discrete logarithm problem. Currently, we're keeping all eggs in one conjectured basket. Even if quantum computing will never be viable, there is a non-zero chance that the discrete logarithm problem will be solved in some other way.
There is a non-zero chance of a black hole forming in the LHC and swallowing the Earth, yet we don’t have a backup plan for that. There are an infinite more examples like that. A non-zero chance is not an argument for anything, unless you have a good basis on actually quantifying it.
