Without going into the details of why nuclear fusion can, and likely would be, a significant, and even necessary, evolution in energy production - suffice to say that it's completion would be a net-positive for the global infrastructure - the above image offers one perspective on why we don't yet have a viable fusion reactor when many of the scientists related to that field have been predicting that it would be a reality "within the next 20 years", for going on well over 50 years!
Too little cheese. A plate with too few clams. Crumbs, but no real cake. Insufficient dough to fill the tummies. Not enough bacon to bring home to the wifey.
Not enough funding.
Apparently, the "fusion scientists in the 1970s" (to quote one of the commentors on the linked reddit thread) projected nuclear fusion to be completed by around 1990 with "maximum effective effort" funding levels, or around $9 billion per year peak funding (blue line on the chart). They projected a 2005 completion with even "moderate" funding.
Yet, here we are in 2018 and we're talking about the potential for a viable nuclear fusion reactor within the next 12 to 20 years and one would think that the significantly less than $1 billion per year funding (black line on the chart) over the course of the last 30+ years is at least partially to blame for this serious gap between the nuclear fusion expectations (back in the 70's) and its current reality.
More on this, and some other interesting facts about nuclear fusion, in the following educational, and surprisingly entertaining, video, courtesy of Joe Scott:
By all means, give the man a follow if ya dig what he's dishin'.