TTD has 64 track combinations, while Simutrans only has 16
This was a decision right from the start of the project. To change that now, means to write Simutrans again - we surely will not do that.
The other side is, that I think the 16 combinations are sufficient to build all kind of networks that you'll need in the game. So even if it's less realistic and less than in TTD I think it's sufficient.
Trains in Simutrans use quite a different routing algorithm than trains in TTD do. They need less guidance by the player. In TTD you use the track combinations to tell the trains where they need to go. Simutrans trains have their own idea where to go, and need less guidance by track constructs. I think that's the reason why Simutrans works well with only 16 track combinations, but TTD needs 64. It's also the reason why TTD players who are used to force trains in a certain direction by clever track constructs just experience problems in Simutrans. In Simutrans the train schedule and one-way signals are the core instruments to guide trains. Switches are vastly irrelevant, the trains know which branch to take.
Also the internal data structures of Simutrans differ quite a lot. That's why two tracks end-to-end connect in TTD while they don't connect in Simutrans.
You can say I made mistakes while designing Simutrans. You are probably right from your point of view as a player.
But I started Simutrans mostly to learn C++ and OOD/OOP. I was a big TTD fan back then (1997), so I used a transport sim as my training project. From that point Simutrans was quite successful. I've learned a lot, not only about program design and coding but also about project management. In that sense Simutrans is a very successful project.
I didn't start the project to create a successor for TTD. Really - back then if you had asked me I'd have said: "impossble".
Simutrans isn't a replacement for TTD and it was most of the time not meant to become one. I know a while I said others, but I've changed my mind - Simutrans currently can't beat TTD in some aspects, and I'm too lazy to change such large areas of it as required.
Anyways, I'm quite surprised how well it evolved and I'm kind of proud what Simutrans became today. If TTD is superior that's ok. But behind TTD only few transport simualtion games can beat Simutrans, and all of them are commercial projects. OpenTTD IMO doesn't count in that competition, because they got a headstart by reusing the TTD code.
I'm very proud that I - being a hobbyist in this regard - could create something that's almost as good as the grandfather and undefeated number one of transport sim games.
If you tell me that TTD has more track combinations than Simutrans has, and that's the reason why you don't play Simutrans, I'll have to accept that.
I've designed Simutrans in a certain manner - mostly with simplicity in mind - and if it is now simpler than TTD in some regard, that's ok for me. E.g. Simutrans was designed with 90 degree geometry in mind. I thought - and still think - diagonal tracks are not needed in a game like this. But players asked me to include diagonal tracks, and I tried to get the best out of the existing structure. Now there are tracks that look diagonal, and with some triockery you can build even long diagonal tracks. But everyone notices that this isn't the thing Simutrans was made for.
User friendliness also was no a priority for me during large stages of the project. It was just a training project for me, so wyh shoudl I care about if other people can use it? If currently the biggest complaint is that Simutrans has "ergonomy problems" that's just because I didn't care much about ergonomy. My goals were different, and the result is, well, quite that what could be expected from my decisions.
Today, I see it from a different point: I see quite a large group of players, and I see that the project isn't just my programming training playground anymore. Most players experience Simutrans as a game in about the same league as TTD, at least they constantly compare it with TTD.
Lately I try to get Simutrans play well in that league, yet I can't undo the decisions of the past. I try to improve the user-friendliness, and I try to expand the features in the game. But the core isn't suitable to replace or supersede TTD in all areas, only in some.
What to do now? I don't know. I still like working on Simutrans, and maybe some day it really will be a competitor for TTD. OpenTTD is a hard challenge. When the TTD source was closed I saw a chance that Simutrans can catch up and supersede TTD, but now that OpenTTD develops quickly, I don't see much chances for Simutrans in that regard.
If OpenTTD was available in 1997, when I was about to start Simutrans, I probably would have joined them. But now I've spend just too much time on Simutrans and I don't want to throw it all away. I'll do what I did the past years: slowly expand Simutrans with new features, and fix the old problems.
So to say: if you have ideas how to improve Simutrans I'll listen. Some things I can't do, some things I don't want to do, and some things I will do.
But if you ask me to let Simutrans become another TTD, I'll say: "no". In that case you're better off playing TTD.
I rather want to develop Simutrans into it's own direction. IMO Simutrans is IMO still too similar to TTD. OTOH it's hard to do a transport sim in a radically different style, so some overlaps will always be there.
'nuff said
And you can build compact two-track mainlines, without these rooms between the tracks. If I cant build such compact two-track lines, I'm not feeling good while building them, destroy them, and build a street, because that don't looks so "wrong" 
