I'm not sure we can extrapolate here: Right now it's expensive government-driven research, Manhattan project style.

Once it's proven that you can make money with fusion (and Wendelstein is our best bet), all hell will break loose.

In programmer speak, you're essentially asking for how long will it take to implement some new feature (let's say collaborative text editor) in language unknown to you :P

Could be months, years, or even decades.

This humorous answer for fusion is 50 years. Always another 50 years.

Realistically, there has been much more progress recently including several SV style startups with massive funding trying to get things to work. It probably is within 50 years this time with the amount of real money and research going into it.