Hello Frontier
Excited to find out that after all these long years, work is being done on a new roller-coaster related game.
The game appears to still be in an early development phase as there is no official information available just yet, perhaps I could make a couple of suggestions.
I thoroughly enjoyed RCT1 & 2, and I thought RCT3 was rather mediocre -- it was a little too slow to my preferences.
The game's pace felt pretty perfect in the original games. Visitors getting in and out of attractions went smooth, whereas in RCT3 it took them a lot longer. I think the "flow" of the game should lean closer to 1 & 2 instead of RCT3.
The idea of all possible kinds of roller coasters being available is fantastic, but my biggest frustration was that there were so many of them that the list got quickly cluttered up, while often a certain coaster would pretty much be the same as an other, except with a different type of train. However, this is a tough problem to overcome.
In reality, for a roller coaster constructor, they start off at a couple of root-coaster-types:
Steel-coaster
Wooden-coaster
Mega-coaster
Hyper-coaster
Launch-coaster
Water-coaster
Powered-coaster
...
The type of track and type of train, as well as the type of elements that are going to be available, is decided from that point on.
If you use a drop-down menu while building a coaster, where you first select your overall coaster-type, next you select your track & train-type, and then you can start constructing it and adding elements that are available for the selection you have made, I think it would work better.
After all, most every type of standard coaster can contain a looping, so there is no need for all kinds of different looping-coasters to clutter your screen. But if you select a Water-coaster, the looping element becomes unavailable, just the same that if you choose a Hyper-coaster, the splash section becomes unavailable, and so forth.
It would keep the game's menu clean and give the player the ability to really create the ride he wants to, as long as it lives up to what's realistically possible.
There are also major differences between coaster-constructors in real life. A B&M uses entirely different track types and trains than a Vekoma or an Intamin. Though I don't know how something like this could be implemented in the game, especially as not a lot of people are aware of the differences, but it could still be of interest to keep in the back of your mind while designing this game.
Of course you couldn't use all of the constructor's names in the game, this would cause way too much licensing issues, but if you could make up your own constructors inspired by the real ones, it could provide a fun extra level of immersion. Am I going to invest in a really expensive, luxurious roller coaster with high quality themeing, or just a run off the mill default coaster by a cheaper constructor? After all, for the price of the high-quality one, I could probably build two less expensive ones, or a coaster and a flat-ride. These provide interesting choices for your parks -- quality vs quantity.
Park management should be number 1 priority, riding the rides yourself should come in second. It was nicely done in RCT3 and a fun feature, but it wasn't what I played the game for. While playing the RCT series, I've only ever wanted to build theme parks that I would like to visit myself, and to see them being succesfully; seeing trains get filled up, selling umbrellas when the rains comes in, watching the people enjoy a good (questionable

) hamburger, seeing my rides all working in their glory, etc. That, to me, is what a fun theme-park game is all about and what makes it fun.
Also, what I've always really missed in the RCT series were proper restaurants. Of course stalls are important in parks, but every park I know has at least 1 actual restaurant. Being able to construct the layout of a restaurant myself and configure the menus and prices would be absolutely brilliant, but I'd already be very happy if there was a list of default restaurants to choose from.
Working with seasons as in reality is another interesting idea, but I don't know how good it could work. This adds another interesting type of choices to make: will my park stay open year-round or will I be open only in the typical themepark season? Will I start constructing a new ride in the fall, to open up in the spring? Though if this system is worked with, game-time days should last longer than they did in the previous games. Seeing your park be quite empty on a cold day could give you the idea to construct a nice indoor-ride or even zone. By the time it's finished, you could start to see a lot more visitors even on colder days. This would be another level of immersion within the game.
I hope these ideas come in useful, and I wish Frontier lots of luck and good times in developing this project, I will follow this one closely.
--krisdw