Tracker modules are great, and I've been listening to them in various forms practically forever. However, the issue is that all of the modules that are desirable are floating around various websites and not always in a ready-to-go format. Even worse, there are some amazing under-rated artists whose modules only exist on a single website and it's not easy to find these artists or their modules without knowing them to begin with!
I've spent probably 300 hours auditing modules to solve this problem. This was caused by:
Limitations with websites disallowing a single massive batch download (or rather, they didn't WANT to provide this option)
Having to discover quality artists in a sea of hundreds of thousands
Fixing any errors in the modules themselves, and in the case of .MODs, renaming them so that the extensions are NOT in the Amiga scheme
Actually having to listen to the batch downloads from each artist to hand pick the good modules from the bad
The sheer fact there's over 48 hours worth of modules in here
Some of the artists that are present in these archives: Alex Kunz, Allister Brimble, Andreas Starr, Andrew Sega, Bjorn Lynne, Dan Grandpre, Daniel Falk, Frederic Motte, Geir Magne Austrheim, Hannu Salonen, Hans van Vliet, Jani Timoskainen, Jeroen Tel, Jogeir Liljedahl, Jussi-Matti Salmela, Matthew Simmonds, Peter Hajba, Rob Gergely, Romeo Knight, Sascha von Dages, Seppo Hurme, Tim Wright, Staffan Melin, Torsten Fassbender-Wellenrausch, Valery Belozercev, Volker Tripp, and many many more...
Unfortunately the quality between .MODs and .S3M or .XM can vary (in terms of sample quality), but if that doesn't bother you I would suggest merging all four of the archives and then playing them in a randomized playlist.
The Amiga .MOD format originally played at 438 Hz due to limitations with the hardware curiously enough, something to do with processor cycles if I recall? This means that most of the time when you're listening to Amiga modules they're not actually running at a 440 Hz pitch which is a partial contributor to the 'Amiga sound'. FastTracker 2 on the other hand was written for MS-DOS PCs and runs the modules at 440 Hz, and also has some minor differences in the way the tracker data is processed.
What this means... is that not all .MODs are created equal. There are some .MODs in this archive which require the FastTracker 2 mode to be interpreted 'correctly', but this can also break non-FastTracker 2 modules, and vice versa!
The good news is, running all of the .MODs in either mode won't really cause that much of a difference except for probably around 30 ~ 40 modules? I've not spent the time to isolate the exact ones due to the sheer size of this, but if it really bothers someone that much (to which most people aren't even AWARE of this anyways) they can go ahead and do the full audit themselves in both modes. Personally, most of the time I listen to modules in the 'FT2' mode, but modules requiring the original Amiga parmeters will sound very strange.
The best demonstration of 'extremes' with the two .MOD interpretations are Jani (C-Quence) Timoskainen's modules which are in fact meant for FastTracker 2 and become distorted very quickly with standard Amiga parameters. I would suggest using balance.MOD and going between the different modes to really hear the difference. If you're using XMPlay you'll need to restart the module each time else weird things happen if you change interpretation modes in real time and don't reset it. I've no idea if Audacious supports the different modes -- I know Stian Sebastian Skjelstad's UNIX port of the Open Cubic Player supports all of the modes and even special flags.