The problem I see is that the measure chosen is a bit of a weird one; it seems to be the ratio between tank casualty rate and total personnel casualty rate, including non-AFV-crew personnel. I find it hard to imagine what this is supposed to measure. If we compared total tank casualties with total personnel casualties, making some allowance for the proportion of personnel in tanks, it might make some sort of sense.
Looking at the US tank bns, and spreading the losses over 300 days, it seems that they had something like one and a bit personnel casualties a day, and a tank casualty every two or three days. They lost personnel at a rate of about three or four per tank. The recce squadron seems to have suffered similar personnel losses per unit time, despite having more people, and with one-third the number of tanks only lost about one every fortnight (insert caveats about means being largely meaningless).
As Connard Sage has pointed out, this is hardly likely to be comparable to the losses for the British units spread out over 3 days (and probably not even that; I expect the losses for 2 Norfolk Yeo were pretty much all on the same day). 8 H and 148 RAC really have too few losses to count, 2 tanks and 1 tank respectively. 2 WG is a recce unit, and so may not be typical, but it has lost 5 tanks per day and slightly less than one man per tank lost (which might reflect the good escape prospects from a Cromwell, or might indicate a greater likelihood for recce tanks to fall victim to mines or obstacles of all sorts in their recce role, or might be just random variation on a tiny sample). The only British armoured unit heavily engaged is 2 Norfolk Yeo, with about 17 personnel and 12 tank casualties per day (and as I said probably really all 50 personnel and 37 tanks lost on a single day). The full orbat figure of 72 tanks is perhaps a bit misleading, too; 11 of these were Honeys, which were regarded as useful for recce, but not combat. At full strength (another questionable assumption) there would have been only 273 personnel and 57 tanks in the fighting echelon of the sabre squadrons, and on that basis the casualties suffered were 18% in personnel and 65% in tanks. 2 Norfolk Yeo’s losses of 1.35 personnel casualties per tank casualty may seem a touch on the low side compared to the campaign average, but it wouldn’t take a vast number of minings or boggings to account for it.
Ivan has mentioned the difficulty of comparing casualties between nations that use different recording conventions. Even assuming the figures given are commensurable, it seems to me deeply improbable that the US figures of 3 or 4 personnel casualties per tank casualty reflect casualties to crewmen fighting in their tanks. The average over the war was more like 2 to 2.5 casualties per tank destroyed, and there’s also the point that a good 20% of the US tank casualties over 300 days were probably mechanical breakdowns, and I can’t imagine the proportion for Goodwood being that high. Three or four crewmen per tank loss is simply not believable. I expect the US casualties include a fair contribution from people in the echelon being hit by German mortars or artillery (or perhaps Allied airpower) which have no connection to the tank losses at all.
All the best,