Where to begin...
I have now been working on my current layout for almost 20 years, and I have been working on the rolling stock and structures on the layout for almost 30 years. If I was to start over again, there are a number of things that I would change, in addition to some that I would keep.
I would stay in HO scale. I have never considered nor been interested in larger scales, and while there have been incredible improvements in N-scale over the last 20 years, my eyesight hasn't been as fortunate.
Hidden staging should never be hidden. In another area where it's readily accessible and easy to get to- yes. Hidden- no. Those lift-outs made from styrofoam panels that I used to access the hidden staging tracks worked great- until they were covered with scenery. And yes, there have been some good articles on building lift-outs, but they are still a pain to use. I would also avoid long runs hidden by anything.
I would use the largest possible radius for my curves. I considered 30" to be more than adequate when I began, and it works well with 6-axle diesels and 50- 60' freight cars. But 85' passenger cars and inter-modal flats just don't look right to me on curves that tight. They might work ok, but the appearance really suffers. I would look to some of the insights in MRH regarding minimum radii for whatever lengths of rolling stock I planned to use.
Hand-laying track and turnouts is something that I have always enjoyed. If I was starting over, I would probably use a mix of commercial track and hand-laid turnouts using CV ties. It would certainly speed up the process of getting the layout up and running, and with a little paint and weathering it can look better than hand-laid track.
I have always enjoyed scratch-building and kit-bashing structures, and there's something about having a structure that no one else has that has always appealed to me. I would continue along this path, even though it takes longer to get scenes completed.
There are a number of people on this site who are building layouts and it's clear that they have spent a lot of time and effort in the planning process. There's a good reason for this, and I suspect that some, if not most, of them have built layouts previously. I jumped in and started working on the bench work as soon as the drop ceiling was in place and the backdrop was complete. I hung a few fluorescent lights around the layout with the thought that I would come back later and install single tube fluorescent around the edge of the bench work behind valances. Installing the single tubes after some of the scenery was installed was a real pain. And I never did get them to line up the way I wanted because of the method in which I hung them from the drop ceiling. And the valances didn't quite cover the lights the way I had intended. There are a lot of good reasons to think these things through before you begin construction of the layout.
I'm not sure if I would stick with proto-freelancing or pick a specific prototype. I have thoroughly enjoyed the creative side of coming up with a plausible location, operational characteristics, time period, paint schemes (both current and past) and all the other attendant pieces associated with a building a believable model of what might pass for a real railroad. But I have also started collecting steam engines and rolling stock of the C&O, and I have long been deeply interested in the coal operations of the C&O and Virginian in West Virginia. Perhaps now that the kids are in college, one of their bedrooms will be confiscated for a brief foray into prototype modeling.
One of the biggest mistakes I made was violating the "pass through each scene once" rule. In order to maximize the length of the main line, the track currently goes to the end of a peninsula, enters a tunnel and then comes back out on the same side of the peninsula while gaining elevation. When the main line gets back to the base of the peninsula, it goes into a tunnel that goes through the backdrop to the other side. You can watch the engines enter the tunnel, but you better make tracks if you want to see the engines come out of the tunnel portal on the other side of the peninsula. If I was starting over, I would seriously consider a double deck layout with shallower scenery in order to avoid this mistake and have a longer main line run.
I've rambled long enough. Suffice it to say that the most important change I would make would be to take very seriously the recommendations and advise of others who have already been where I plan to go. In a lot of instances, I thought I knew better, or that my design/concept/idea would work out just fine in spite of advise to the contrary.
Tom Patterson