What is interesting about this article is that it was published about a week before the publication of the results that show that there may be an error in the radiocarbon calibration curve in the 16th and 17th centuries BC. So in this respect, at the time "The Atlantic" was writing, the Olive branch was indeed the contentious evidence, but now, with the new proposed radiocarbon calibration curve, its true calendar age will be younger and so more in line with archaeological evidence.
There is still of course the issue of growth problems with Olive, which makes identifying annual rings difficult, as well as missing annual rings entirely, but this can be somewhat nullified (at the expense of some precision and accuracy) by ignoring all tree ring information and treating the sample as a series of segments (rather than a series of accurately counted tree rings) and calibrating the radiocarbon ages using an ordered sequence analysis.
The real issue though is that the radiocarbon ages from artefacts directly associated with the eruption of Thera now lie upon a radiocarbon plateau (a region of the calibration curve that is essentially flat) on the new proposed calibration curve. What this means is that although radiocarbon dates may now be more accurate, they will not be as precise. So for example, if we take the wiggle matched radiocarbon date for the Olive branch and assume that the annual rings are accurately counted and there are no missing rings, then using the current calibration curve IntCal13 we get a date range between 1626-1605 BC, a range of only 21 years. With the new proposed curve, this range becomes 1610-1567 BC, a range of 43 years. It is likely more accurate, but it is half as precise.
Jonny
The path to good scholarship is paved with imagined patterns. - David M Raup