Great response. I think a combination of two strategies should be considered: main HVDC trunk lines for concentrated demand sites (server farms, cities) and then localized/off-grid systems with storage for small towns and individual users.
Solar generation isn't exactly "free", but sunlight is. High demand sites require power to be fed from somewhere unless they build local (fusion, geothermal, gas-fired) plants as part of infrastructure. As solar gets better (perovskites, etc) and batteries get better, both battery and solar-produced hydrogen backups become viable for mid-size applications, but the top-end users are still going to need feed lines. HVDC is the most efficient for that, now that electronic switching has become reliable enough to produce and transform it.
If China holds the cards for buildup, then she should be encouraged to spend on the backbone and get rewarded for it.
I'm skeptical of any 'expert' that says something isn't viable; especially when it's a university or an unthink-tank.
Investment is fickle at best because stock markets have been blowing up uselessness for so long.
The paper by Downie appears to be bogged down in political anachronisms of competitive failures to trade power. GEI is about cooperation and need, not prices. That's why contemporary analysts hate it. As climate disrupts weather, population geography and food supplies, more and more countries are going to be trading necessities rather than trinkets and baubles. A global solar grid will look better to investors and politicians every year.