Yes, SMR's main focus at present is on coking coal from Isaac...

Yes, SMR's main focus at present is on coking coal from Isaac Plains. But, from the company website: "Stanmore continues to progress its prospective high quality thermal coal assets in the Northern Surat Basin which will prove to be valuable as the demand for high quality, low impurity thermal coal grows at a global level.". So sharply higher thermal coal use by China can't exactly hurt.
As for solar being cheaper than coal/gas/oil/uranium by 2020 - well, it might be, but I wouldn't hold your breath while waiting. [Warning: What follows is is a US-based study http://instituteforenergyresearch.org/studies/levelized-cost-of-new-generating-technologies/]. LCOE (Levelised Cost Of Energy] in 2018 is estimated as follows -
Conventional coal - 100.1
Natgas - 67.1, 65.6 (CCC, ACC)
Nuclear - 108.4
Solar PV - 144.3
That gives Solar PV a lot of catching up to do from 2018 to 2020. But the above data is only a part of the story. What proponents of unreliable energy constantly ignore is ... the cost of unreliability (as seen rather dramatically in South Australia recently). The report says:
"What EIA is expressing is that dispatchable technology costs should not be compared to non-dispatchable technology costs because the latter technologies only supply electricity generation when the resource (e.g. wind or sun) is available, but they do not supply capacity that can be relied on to provide electricity. IER reported on one analysis that attempts to measure the “levelized avoided cost” of wind, for example. In this paper, the hidden costs of wind (e.g. the cost of back-up power) added to the levelized cost of wind totals 15.1 cents per kilowatt-hour if natural gas is used as the back-up power and 19.2 cents per kilowatt-hour if coal is used as the back-up power.
It is important that readers, especially policy makers, understand this aspect of non-dispatchable power. Since non-dispatchable power cannot be counted on to produce power when the consumer needs it, it is, in a sense, an unconventional electricity source. Because our electrical system must respond to consumer demand instantaneously, non-dispatchable power is in essence superfluous to our needs. The requirement that dispatchable power back-up non-dispatchable power to make sure electricity is there when needed is not a luxury, but a necessity. The more that non-dispatchable power is used, the more the electrical system requires investments in dispatchable generation forms to back up its increased use. Government policies that promote the use of non-dispatchable power are equivalent to requiring consumers to buy and care for two vehicles: one that works when you need it and another that works when it feels like it. The hidden costs of non-dispatchable power are substantial and should not be overlooked as part of the public policy discussion."

To my simple mind, if people didn't have to use unreliable power (MRET etc) then they wouldn't use it now, and they wouldn't use it in 2018 or 2020 or whenever. As they say above, unreliable power is in essence superfluous to our needs.