I've done a few tests, and in my case, it does seem to be a "you're holding it wrong" issue.
I used an airport extreme base station (before the dual band one) running in b/g compatible mode, 2.4 GHz, and the latest firmware for it.
Also MBP, MB air, iphone 4 and new ipad 32/wifi only.
Our far bedroom in our wooden, single story house is the challenging wifi spot--we knew this already.
For those of you with Airport Extreme base stations, you can use airport utility (except, unfortunately, the newest version--I used 5.6) to monitor signal and noise strength. This is under Advanced, Logs & Stats button, wireless clients tab.
I monitored signal strength (noise didn't change) for 4 devices.
In general, the two laptops had a few decibel better signal than the phone or ipad.
If propped up on the bed in any orientation, the ipad had about the same signal as the nearby iphone 4. Also, about the same connection rate and speedtest results, with the portrait orientation giving 1 or 2 decibel lower signals.
However, if I had the ipad on my lap, then I could cause a loss of signal by holding the bottom corners in portrait mode. This could cause a loss of about 8-12 decibels, and enough to degrade or lose the throughput.
I did not see this in landscape, although landscape button right seemed to be a bit better than landscape button left.
If I propped in in my lap, in portrait, but didn't hold it, I got the same results as when it was propped on the bed by itself.
So in general, the new ipad performed about the same as the iphone 4 except when held at the bottom in portrait mode, when the signal would degrade about 10 db.
I probably won't be reporting this as an applecare issue, as we need to improve the signal in that part of the house anyway, and I can switch to portrait if I'm having trouble connect somewhere.
I don't have a case, so can't report on that.
Ron