A replacement aftermarket filter element (using the OEM housing and plumbing) is unlikely to have ANY effect on MPG or performance. You can count only on gaining the convenience of being able to clean the filter rather than having to replace it.... which you may not think is such a big concern after you read the rest of this.
Most people change their filters too often. If you read the EPA test referenced in the link below, you will see that even severe reductions in airflow from plugged filters has little to do with fuel economy because the EFI system continually trims the mixture. Even carbureted cars are less effected than many of us might think.
Air Fitlers and Fuel Economy
The gist is that a filter looking a little dirty is likely still doing a fine job at filtering and is likely to still have all the airflow capacity the engine needs. It likely will continue good filtering right up to the point it's completely plugged and the engine won't run above idle. Filtering efficiency actually improves as the filter loads up a little but the potential danger in a dirty filter is that if the restriction gets severe enough, the filter element may fail and allow the engine to start sucking dirt through.
If you have a restriction gauge, you can forestall that possibility. The Donaldson Informer or Filter Minder (you can Google them) are two easily available restriction gauges. The general rule of thumb is to replace a filter when the restriction increases 2.5 kPa (0.75" mercury) over the stock clean filter reading. Pulling filter elements out often for cleaning or inspection is often counter productive because you can ruin the filter sealing or drop dirt into the intake.
How plugged can they be and still allow the engine to perform? I have done some flow bench testing on filters and many of the OEM elements have more than enough potential airflow to allow the engine to produce maximum power, even when partially restricted and dirty. Case in point, I recently flow tested my stock 5.4L F150 filter with it's OE Motorcraft filter element in it (w/20K miles on it) and it delivered an average airflow of 621.58 CFM in the stock housing. The calculated airflow needs for the 5.4L engine at 5200 rpm is around 400 cfm (depending on what volumetric efficiency you plug into the calculation... I used 0.80). That means I have 35+ more airflow than I need for maximum power WITH A FILTER THAT ALREADY HAS 20K MILES ON IT! I can't speak to every filter or vehicle, but the generic airflow figure I have seen is that most OE filters have 40 percent more calculated flow than the engine can use to account for dirt loading. I seldom see 5200 rpm (the truck sucks enough gas as it is!!) and feather-footing it around at 2000 rpm and using that prodigious low end torgue to haul me around, the airflow needs are more like 150 cfm, so at that rate, I have more like 75 percent more airflow than I can use at that moment.
There is no power to be gained just by having more potential airflow capability than the engine can use. As mentioned earlier, the overage in the stock filters is to account for dirt loads. The engine needs what it needs at the moment and, if it gets it, other factors will hinder performance, not airflow. When it runs out of air, that's when you'll see a drop. By itself, on a stock engine, a complete system that is well tuned (runner length and diameter, cool air ducting etc.) for the car may have an effect on performance at the higher RPM range, perhaps on the midrange and perhaps a very small effect on MPG (though not likely). How much gain is proportional to how well the OE system was designed vs the new one. The OEMs are doing a pretty good job designing intake systems, spending a lot of time and money on them to wring the most power/MPG and, frankly, some of the aftermarket suppliers do little if any R&D or tuning... just making something that looks good and has gnarly intake noise.
When you have increased the airflow needs of the engine by other mods, that when you can outstrip the OE intake and replacing it with a higher flowing system is necessary more to NOT LOSE the power from the other mods by restricting airflow than it is gaining power from the airflow increase itself.
As mentioned, the filtering ability of some aftermarket media is questionable... just as the quality of some aftermarket replacement filters are questionable. There hasn't been near enough study done IMO. The OE and the better name brands are a surer bet towards meeting the original design specs than some of the super cheap auto parts store, of no-name house brand filters.