Which is why we need gain of function experiments to accelerate this process and get ahead of the virus so that we are prepared for the eventual successful mutation.
If you have labs picking-and-choosing what mutations to make (what usually happens), then no. That's just hubris. People stumble around in the dark, occasionally find a truffle, and pat themselves on the back for having such a great truffle-finding method.
Even for good artificial selection systems, mutational space is so gigantic that it's hard to cover properly: not just the point mutations (i.e. converting one amino acid to another) but also insertions, deletions and transpositions. There's no artificial selection system I'm aware of that can recapitulate mutational space.
For example we can try spikes from different related viruses, or pick and choose various other tricks, like furin cleavage sites, for example like in this rejected 2018 EcoHealth coronavirus research DARPA grant proposal: