Sure. Depending on the customer, replacing everything with new hardware will vary from a financial impossibility to an invitation to find another consultant. Nothing like a "change everything" solution to a solving a minor problem. Hmmm.... didn't we just elect a president on that principle? Oh-oh.
Besides, technology offers a possible solution. If you can't coexist with them, jam them with more RF:
Anyway, let's get real for just a moment. There's no way you can get a free lunch with wireless. If you want speed, you're going to lose range. If you want reliability, it will cost you both speed and range. If you want backwards compatibility, you have to give it air time which costs speed. Free lunches only exist among congress critters. You have only to look at the product test results on SmallNetBuilder.com to see the distinction between advertised claims and reality.
Even MIMO doesn't offer a free lunch. Since multiple streams depend completely on the presence of reflections, if anything changes to affect those reflections, the thruput falls apart. I've seen this in my "testing" (i.e. screwing around) with an Apple MacBook and Airport Extreme (802.11n) base. While running continuous Iperf or Jperf benchmark tests, the thruput varied all over the place. Thruput would sometimes hit 60Mbits/sec followed by an immediate drop to perhaps
12Mbits/sec. I only had to move the laptop slightly, and everything would change. Watching an H.264 QT video, while walking around the room, was ummmm.... painful. More:With just the Apple stuff running in the room, performance was respectable and methinks impressively high. However, when I fired up a wireless peer to peer network between two other laptops nearby, thruput hit bottom at perhaps 3Mbits/sec.