These are all net returns to the investor with the data the absolute performance number. You'll note that the benchmark is referenced at the bottom. So all those in green are those above benchmark.
Generally speaking, being capped out is a drag because many managers get alpha from the smaller end of their portfolios, it's easier to add value there because there is comparatively greater informational advantages to be had. Obviously when you have dozens of market analysts covering larger stocks, they tend to be price to perfection. So as they get too big, they get less able to invest at that smaller end to any meaningful extent, so the returns fall.
That said, a small number of manager do seem to regularly add alpha even though they are managing too much money. It is a bit difficult to say why this happens, except that they obviously rely less on the smaller end for attribution. Basically it just means that their stockpicking at the higher end is better than peers. Ophir is obviously a case in point. You could say the same about BT. But it's also important to differentiate between those that are closed because they see it as prudent that they don't get too big, and those that are closed (or not) because they are to big.
Regarding significant wholesale investors and fees, this refers to institutional investors such as super funds, sovereign wealth funds etc. This is an accepted practice across the industry and not something I have an issue with. If someone is going to drop in $100 million in one hit, they are probably going to press for a discount. Even cheap index funds charge less for institutional clients.
