You got caught by those silly English units.
Your expression, $I_t = I_{sp}g_0 \dot m \Delta t$, works just fine when you use metric units. The mass flow rate, in metric units, is 127 kg/s. The total impulse, in metric units, is
$$I_t = (217.5\,\text{s})(9.80665\,\text{m}/\text{s}^2)(127\, \text{kg}/\text{s})(65\,\text{s}) = 1.76\times 10^7\,\text{N}\,\text{s}$$
Converting this to customary units yields $3.96\times 10^6\,\text{lbf}\,\text{s}$.
Doing the exact same calculation, this time using customary units, yields
$$I_t = (217.5\,\text{s})(32.17405\,\text{ft}/\text{s}^2)(280\, \text{lbm}/\text{s})(65\,\text{s}) = 1.2736\times 10^8\,\text{but in what units?}$$
The dimensional analysis says the units of this expression are lbm·ft/s. Those aren't the desired units; you want lbf·s. The metric F=ma becomes F=kma in customary units. Dividing by the numerical value of g0 converts lbm·ft/s2 to lbf. That same division here converts lbm·ft/s to lbf·s. And indeed, 1.2736×108/32.17405 is 3.96×106.
This suggests an alternate expression, $I_t = I_{sp} \dot m \Delta t$, where $I_{sp}$ is in seconds.This yields a numerical value of 217.5·280·65=3.96×106. That's the right value (at least numerically). Note well: This alternate expression does not work in metric units. In metric, this calculation yields a value of 1.80×106 kg·s. The correct value is 1.76×107 N·s.
Strictly speaking, the above does not have the correct units; it has units of mass*time. Numerically, it yields the right value in customary units because dividing by the numerical value of g0 cancels the explicit use of g0 in the numerator.
This calculation is not correct in metric units. To get the right metric units from this calculation one needs to multiply the numerical value of that result (1.80×106 kg·s) by the numerical value of g0. Indeed, 1.80×106 * 9.80665 = 1.76×107.