There is a lot you don't know about the structure of the game's files. Each expansion adds new files with the same name as in the original game. These files are used instead of the files in the base game. The files in the base game are ignored. You are using the files in the base game, which is why you can't find "train" or "legion" or "trainlegion". They were added in Crusade and don't appear in any of the base game files.
You are looking in the Galactic Civilizations III directory, which is the correct place to start. After that, you need to look in the correct folder in that directory. The data for the base game is in the data folder. In there are folders like Core, English and Game. All the data for the base game will be in those folders and other folders.
The Crusade expansion adds new data to the existing data or completely replaces the existing data. To do this, new files are used. All new files are in the DLC folder in the Galactic Civilizations III directory. In your case, you will have the folder EXP2_Crusade in the DLC folder. The EXP2_Crusade folder will contain Core, English, Game and other folders that contain the new data for Crusade. You must use the new data for the Crusade expansion. You will be able to find "trainlegion" in the ImprovementDefs.xml file in the Game directory in the EXP2_Crusade directory. Look through the various directories/folders to see what is there. Always look first in the directory for the latest expansion that you have in order to find the file you want. If it's not there, look in the next-oldest expansion, then in the even older one until you get to the first one. If you still haven't found it, use the one for the base game. Sometimes something new will be added in a DLC, so what you need will be in appropriate DLCxx folder in the main DLC folder.
I have all the expansions, and there are multiple versions of some of the files. Using the wrong one will cause problems. I know this from using the wrong one too many times and trying to figure out why things weren't working right.
The files in the various data directories are generic when it comes to names because the game is available in several languages. There has to be way to get the correct name to be used in the languages. This is done by having a language directory, such at the English one mentioned earlier. It will contain a folder named Text. All the conversions from the generic names to the correct name for a language are in the Text folder. The files in there usually have names similar to those in the data directory, but with Text add at the end. In your case, the ImprovementDefs.xml in the Data directory for the Crusade expansion will be ImprovementText.xml in the English\Text directory.for the Crusade expansion.
The first improvement in the "Defs" file is the colony capital. It will contain these lines:
<DisplayName>ColonyCapital_Name</DisplayName><ShortDescription>ColonyCapital_ShortDec</ShortDescription>
The "Text" file will contain these lines:
<StringTable>
<Label>ColonyCapital_Name</Label>
<String>Colony Capital</String>
</StringTable>
<StringTable>
<Label>ColonyCapital_ShortDec</Label>
<String>Provides the basic needs of a colony. </String>
</StringTable>
If the language is not English, the words "Colony Capital" and "Provides the basic needs of a colony." will be different.
Study the various directories and files in them along with the contents of the files to learn how everything is set up and how everything works together. It will probably be difficult and confusing at first, but things will become simpler and more obvious as you learn more and become more experienced.
You should also learn to use search like I did in the screenshot in Reply #1 to look for file names, partial file names and file contents. It can save a lot of time and trouble.