Nobody, not even the team directly responsible, believe that R100 is literally achievable within that time scale. The question is simply how to present reaching 95% or 98% superfast as "meeting" the target without too much flak - maybe lots of satellite dishes. The UK Government's response to BT's USO proposal pretty much scuppers any alternative to BT as the primary R100 contractor as there is little commercial attraction to firm XYZ in offering 30 Mbps via some new network when BT seems likely to provide 10 Mbps on demand via copper. So what is left is all the properties for which the USO cost cap applies.
Even using TV whitespace that is hardly likely to be a viable market and there is no guarantee that the spectrum will be available in time or at reasonable cost. Experimental systems relying upon TV whitespace are not convincing for broadband coverage in remote rural areas. The equipment is far too expensive, channel sizes and cell coverage are too small for high speeds, NLOS is irrelevant when there are hills in the way, etc. Not vapourware but vastly oversold as a mass market product. In practical terms it is much easier to upgrade mobile broadband networks - and no-one believes that is the solution to R100.