A reactor will run for 400,000 ticks. 1 uranium cell with 4 reflectors will produce 25 EU/t. Therefore 10M EU in total. Which would make a profit, although you'd overall be making a loss as you'd be burning up copper, tin and coal in the process.
If you went for a uranium only design, say 100 EU/t for 6 cells you could produce 40M EU, which split between 6 rods would net a theoretical profit of 9M EU, but that's also implying you've got lucky with the scrap. That 9M EU is split over 400,000 ticks, so the actual profit is a mere 22.5EU/t, which is a little more than a geothermal generator.
So apparently yes you could just about squeeze a profit out of Uranium-UU looping, although the actual EU you'd get to use would be rather uncertain until the next set of fuel rods were made, not to mention you'd have to have a mass fab and a replicator dedicated to make uranium fuel rods, which would be burning iron up (unless you UU'd that as well, which with an iron ingot costs 1.066 mb, so another 166,666 EU + scrap + rolling machine x 2). The maths in terms of replication checks out as being right, but I can only think that's because native IC2 is quite generous with Uranium Ore so the uranium values are lower than what they could be. Just increasing them a little could tip it to only making a profit with incredibly efficient uranium designs, but still with the constraint of most of the produced power being used.