Not sure I can use a public gym the same after reading that.
IMO, there is only so much chewing a person is going to do.
Because a person at 200 lb is eating enough calories to support a weight of 300 lb doesn't mean that when they reach 250, they are going to be eating enough to support 350 lb. Thus, the method of eating ALSO doesn't linearly increase with the weight gained. Therefore, it is logical that weight gain will eventually "level off"..
Think of it this way.
Drive 30 mph for an hour... Average speed... 30 mph.
For the next hour, drive at 70 mph. Average over the 2 hours.. 50 mph ... big gain
For the next hour, drive at 70 mph. Average over the 3 hours.. 56 mph... large gain, but not as much
so on ... Average over 4 hours.. 60 mph ... smaller gain
Average over 5 hours... 62 mph ... even smaller gain
average over 6 hours.. 63 mph.... is this a measurable gain?
So, even though I have been driving at 70 mph for most of the time looked at, it appears that I am "leveling off" at 63 mph... which is literally 10% less than my actual speed. What is happening is that I am asymptotically approaching my ACTUAL speed... thus, as time goes along, my rate at which I approach my speed goes WAY down.
Personally, I see weight gain the same way. Given that the calories in are relatively fixed... weight gain is going to asymptotically approach the actual weight for that amount of calories given infinite time. I feel that is why we are seeing the "leveling out".. How much we are eating isn't necessarily increasing that much.. and we have been doing so for long enough now that when we look at the "recent" short term trends, we don't see much change anymore... and we declare the problem fixed!