First locomotive finally ran a full loop around the layout

I ran into a similar wall when I was setting up my power distribution for the siding. Even with the bus under the benchwork, routing all the individual dropper wires cleanly took way longer than I thought it whould. I ended up using a small rail section with pre-drilled holes just to manage the feed lines into the siding turnout rather than trying to solder every single connection directly to the rails. It kept the overhead messy while the wiring was still a bit of a tangle. I was using a mix of 2 and 4 rails for my main block and found that having a power drop every few feet really helped keep the voltage consistent across the whole length. The motors would sometimes twitch if the current dropped at the far end of the loop.

Regarding the tight curves, I had the same trouble pulling wheels off the rails on the inner side. I ended up adding a half inch of extra radius to the inner rail piece, which felt like overkill, but it really smoothed out the contact for the smaller dies I was running. Getting that first successful long run is a huge feeling. I reccomend testing each section with a single train before you commit to the permanent setup, just to catch those wierd spots.

When I was testing out a loop layout last spring, I ran into issues that felt less like a physical stall and more like a power delivery problem. Even though my track sections fit fine and the curves looked perfect on the benchwork, the older N scale locomotives I was using would suddenly start dragging or slow down dramatically after a few seconds of running. It wasn't that the track was too tight, but the current draw seemed to drop off unexpectedly. I eventually traced it back to the distance between the main power feed and the furthest section of the loop.

It turned out that even with a decent bus running under the bench, the voltage drop across such a long run was affecting the motors, causing that sluggish behavior. I ended up running a dedicated booster line, a simple thin wire run closer to the remote end of the layout, feeding a small, low-amperage power module right where the current was failing. It stabilized the speed across the entire run, making the entire layout feel much more consistent. It taught me that the physical track geometry is only half the battle when you are pushing running quality. Getting those curves perfect is satisfying, but making sure the electrical supply keeps up with the operational needs of the entire system is the real work. It's a pain to run those extra power drops, but it makes all the difference when you are trying to simulate a high-speed run on a larger scale.

That last part hit home. I've still got a section of helix sitting on foam blocks with zero fasteners and it's been running fine for a year, so the 'leave it loose till you're sure' approach is legit. One thing I'd say about ballast, don't rush it even after the curve swap. I glued some down too early, then wanted to reroute a turnout and had to chip the whole mess up. Now I run trains for a solid month before anything gets permanently attached. Are you going DCC eventually or sticking with DC? That motor hum sounds diferent on each and I'm curious which you're hearing.

That quiet house hum thing is real. I put a cheap amp speaker under the benchwork a while back and now I leave the layout running just for the background noise while I work on other parts of it. One tip from my own track laying days, mark every rail joint with a paint pen before you glue anything down. When you come back months later to add feeders or fix a kink you will thank yourself for knowing exactly where the seams are without crawling under there with a flashlight.

Gotta push back on the inner rail thing from earlier. Widening just one rail on a curve messes with your gauge, and if it's off the wheels will find every bad spot instead of fewer. When I hit the same problem on my helix approach it turned out the rail joints had crept out of alignment, not the radius itself. I eased the joints and rechecked the gauge with a feeler and the drops stopped. If a curve is truly too sharp, swap the whole section like the OP is doing, don't try to fudge one rail inward. That half inch trick sounds like it's hiding a different fault.