The problem with that system is you don't know where the info comes from or how recent it is ,,
Every rating system will have some problems, it is just a matter or how big they are, and how many there are. In the case of the Fargo ratings I believe they are just about as few and small as could ever be possible and I am sure that if he discovers a way to tweak it to make it even slightly better in the future he would do so although it was exceptionally well thought out and done from the get go and they may be little if anything to ever improve on that doesn't come with a worse trade off. If you believe a better system can be designed, I think you should do it, or at least give your specific inputs along with the solutions in one of the recent threads specifically addressing the details of the system. I am certain that Mike Page is open to all good ideas if they would actually be an improvement with all things considered.
As far as where the info comes from, some educated deduction says that that much of it came straight from CSI, and he probably has arrangements with some other promoters and streamers to provide the statistics as well. He may even use a few trusted people such as AtLarge who track tournament and/or challenge match statistics independently. Are you insinuating that he is making up the match information in his system database and that it isn't legitimate?
As far as how recent his information is, we do know that in the sense that his rating formula takes that into consideration and adjusts for it. New information (recent matches) count the most heavily towards the player rating and the older the matches are the less weight they carry and the less effect they have on your rating. We know that he uses matches from today all the way back to ten years ago. Not sure what else you need to know because the formula takes care of accounting for the "age of the info".
You can't use only the most recent matches for a player for several reasons, but primarily it is just too small of a sample size in many cases. What if a player was in a slump the last few months. It is temporary and they will come out of it at any moment but their rating is going to show them being much worse than they actually are because of it if since you weren't also taking into consideration how they truely played for years just prior to this temporary slump. Plus, many players don't play a billion matches every year like say SVB. If you only play a tournament or two a year, your rating based on say just the last year won't at all be accurate if it just so happened you had an off day in both those tournaments or had people running sets out on you to put you two and out etc. He went for the best compromise and all around most accurate way to do it which was to count all of your matches (that he was able to get the results for) from the last ten years but the recent ones affect your rating a whole lot more than the oldest ones.
And remember, this system was designed to be the all around best when you have to rate everyone in the world, including the nobodies like you and I, and not just the pros that are actively playing professional events. This is another reason you can't just use say only the most recent year or some short period of time. Most people just don't play enough big tournament matches a year for the rating to have much meaning (small sample size). But on the flip side, obviously how you are playing this year does have more meaning than how you were playing say 8 years ago. And that is why Fargo takes it all into consideration for the best of both worlds, using all available matches up to ten years old, to give a bigger more meaningful sample size, but at the same time gearing it to be a more accurate reflection of your current play by having your more recent matches count much more towards your rating.