So, the conclusion could be: forget the delta loop. But that is a dangerous thing to do. First, you have to consider the pattern of spots hearing both antennas. Here's the plot, delta in red, vertical in green:
Clearly, the conclusion has to be modified. The reason the vertical initially appears to be better than the delta loop is that the stronger (sometimes much stronger) reports are coming from close in, within Europe. From the overall pattern, it's clear that the delta loop is reaching much further into the eastern US than the vertical. This plot shows how that works out in numbers:
But note the exception: the highlighted receiver in the map of spots shows that KB1VC heard both antennas at the same strength at the same time. From that, you would expect a few more receivers to have heard the vertical in the US, but they didn't.
The overall conclusion is clear enough: the delta loop (which, as effectively a pair of closely-spaced, phased verticals) should enjoy roughly a 3dB advantage over the 1/4 wave vertical, is indeed the much better antenna, enjoying a significantly lower radiation pattern (as indicated by the weaker short-skip reports) where that gain is realised and turned into DX spots. Sloping ground only means an even better low-angle performance.
The modelled pattern below shows clearly the suppression of high angle signals for a delta loop:
And here's the vertical's pattern, with a nice (but mostly undesirable) high angle bulge. Note how different it is to all those lovely, symmetric patterns you see in books and magazines, who invariably use misleading 'free space' settings (and, sometimes, infeasible radial numbers) to do their computations.
The next test will be placing the vertical at the water's edge. With winter coming, that may have to wait for a while!