It would surely help, but then, having all antennas close to each other isn't very good, due to unpredictable results of so-called radiation coupling. It may, or may not be severe, depending on random conditions - hard/impossible to predict, so they're put on different sides of device - basically, as far from each other-, as possible/feasible - for a purpose. Note, that even N950 put them to both sides (thus complicating body design, due to two plastic parts above antenna, instead of one).

Coupling mentioned above would happen no matter how many antennas are actively used at the same time. But, there is other possible problem - active interference, when using radios simultaneously. Thankfully, we're rarely using both 3G and WiFi transmission at the same time (3G is closer frequency to WiFi, while GSM/2G is far away, much lower), but to minimize it, users would need to get habit of switching to 2G-only, while on WiFi. I do it intuitively for battery saving purposes, but for some it may be PITA to remember.

Now, it is possible that you may just scratch all of above, as you would never encounter any problems in real life. sadly, no other way to test, than doing prototype and test-driving it (and adjust in case of failure, then re-do prototype, then test drive again... and again...). Despite advancements on antenna design simulators, such complicated array of antennas and feeds so close together - with half of feeds hardware properties unknown/closed - is simply uncomputable.

That's why guys designing mobiles are doing it first, together with whole board project. We meet existing design, trying to put it in different "box", (and/or using antenna replacements) changing layout, etc - it's *much* more complicated than some would think.

Funny enough, the design that should work in *all* cases, would be to add stick-type antenna, like in those early days mobiles:



Putting antenna outside "hull", far away from noises, interferences, heat, etc, clear 99,99% of problems.

/Estel