Yes, the very densest materials should settle 'at the bottom', which in this case is the very center of the earth's core. Uranium and Gold have similar densities (about 19), but I suspect there is more Uranium. In any event, there should be some occasional fission of both isotopes (U-238 and U-235), with the U-235 happening more frequently. Whether this produces more energy, or significantly less energy, than the
radioactive decay chain energy production, which is well known and independent of the amount of material, is the question.
On our planet's surface, Uranium is never present in amounts greater than a few meters in thickness. Since the neutrons from fission (both lesser amounts of U-238 and far greater amounts of U-235) are very fast when produced and not very good at fissioning Uranium atoms, they need to slow way down in order to be good at fissioning. (This might seem counter-intuitive to some, but in nuclear parlance, the
cross-section for fission increases as the speed of the fissioning neutron decreases.)
On our surface, to get around this obstacle and establish a continuous fission reaction, we add materials that will make the speeds of the neutrons more moderate (slower, down to thermal speeds), called
moderators. Water is a common one, and can double as a heat-extractor. Carbon also works.
In the center of the earth, if there is a large bolus of Uranium, simply the long distance from side to side of the bolus (miles?) could serve to moderate fast neutrons into slow ones. But it is not known if there is a large abundance of
poisons in the bolus that would greatly slow or stop the reaction completely.