If she likes it, she can afford it, and it's performing well, I'd keep it. If it's stretching her budget or she can get by on a much cheaper computer, return it.
There's a saying:
Correlation does not imply causation.
If you already know that bit of wisdom, it's a good time to think about what it means. If you're not sure what it means, I'll apply it to Flexgate.
We know there appear to be a significant number of earlier model tbMBPs that have this "spotlight lighting" issue with their displays that eventually result in screen failures. When those machines were opened up, it was discovered that the cables appeared to be too short to handle the stress of opening the lid all the way. Therefore we conclude that the length of the cable was the cause.
What's wrong with this conclusion?
Most 2016 and 2017 tbMBPs that don't have this issue probably have the same cable length.
Apple's certainly has stress tested their laptops during development. They have robots that will open the lid more times in a day than you'll ever do over the lifespan of long lived laptop. Simple risk factors like ordinary mechanical stress alone would likely have been identified during development. If the cable length is indeed as problematic as some believe, it's likely that it's not just the cable length and opening of the lid absent of other factors that cause Flexgate. For all we know, it could be due to bad cables sneaking through the supply chain, which would mean that there are good cables and bad cables floating around out there that are identical in length and appearance.
People who say things like
"it's only a matter of time before X fails" with only circumstantial browsing of the Internet to state their case violate the basic rule of statistical analysis above. I'm not saying that they're wrong to believe and I'm right to be skeptical, but I am saying they lack the basis to be so confident.
If you're still not sure what I mean, here are some humorous examples of correlation and causation gone wrong.
http://www.tylervigen.com/spurious-correlations
I had a 2016 MBP for 29 months that was used daily without precaution and often with the lid open to the max. I didn't have flexgate. I didn't have serious keyboard issues either. With the exception of one stiff arrow key that was stiff from day 1, my keyboard worked fine.
If you're going to get your wisdom from MacRumors and similar sites, repeat after me:
Correlation does not imply causation.