I'd be careful of how close you place those radiocarbon dates.
Radiocarbon dates are not generally expressed as single figures (intercepts) but rather as ranges. Those ranges can either be 1 sigma (67% chance of being within that range) or 2 sigma (95% chance of being within that range). 2 sigma is the standard for reporting.
So, a date might be expressed as cal 2 sigma 1890-1630 BC (3840-3580 BP). The intercepts for this particular date are 1750 BC and 3700 BP, but those are not necessarily accurate. The range is the figure that is statistically accurate. And that's a pretty tight date, with a ±30 range.
But, given a large number of dates on closely related materials that begins to change, and you can get much tighter dating if things work out right. It sounds like that is what may have been done in this case.
By the way, in my work as an archaeologist (not in the Old World though) I have received 632 radiocarbon dates, have six out being processed now, and have one more going out tomorrow.