Consider this: at a time when AMD have finally found some financial stability and starting to regain technological and performance leadership, do you think they are desperate enough to lose money on Vega 20 packages just to achieve some superficial PR win and some kind of parity with Nvidia?
I would only say "yes" if many factors fell into alignment.
Personally, I think Radeon 7 is powered by a Vega 20 package that didn't meet the requirements in testing to be a MI50, more than likely power draw is too high. So that package is scheduled to be discarded, and AMD take the financial hit. It's only by virtue of Turing's raster performance being underwhelming and astronomical prices that a window has opened up for AMD to repurpose those trash-destined MI50 packages as a gaming card and recoup some kind of money from it; gaming cards can afford to have their power draw and TDP ramped up, so AMD did just that (clocks too), slapped it onto a PCB that was already designed, mated it with a cooler that was already designed, and lo the Radeon 7 was born.
If Turing saw significant raster gains over Pascal, I don't think AMD would have bothered with Radeon 7 because they still couldn't compete
If RTX cards were couple hundred bucks cheaper, I don't think AMD would have bothered with Radeon 7 because they couldn't match prices
Of course we'll never know what AMD's thought process is regarding Radeon 7, but for me it makes no sense to give up thousands of dollars per sale of MI50 cards to divert their Vega packages to create a Radeon 7 as an actual, proper and planned graphics card. The end result is Rayfield's "unfeasible" has actually come into existence, but I believe it's from aligning circumstance, not by design.