As an OldPilot with all your ratings I'm sure you've got thousands of IMC hours but how much of it was flown using ADS-B IN and a portable app? Let's keep it apples to apples. Y'all keep pointing out that the blocky returns are "accurate" and I keep pointing out that they're not, and I give examples of why I think they're not, and no one has yet offered evidence or examples that I'm wrong. So the important question to be settled is, how accurate is the current ADS-B return? I just downloaded nexrad on iFly (internet connection). Brian has mentioned above "Currently we do want to keep the Internet and ADS-B consistent", so the ADS-B download should look the same as the internet download I'm looking at. I searched around and found a block bordered by recognizable landmarks (shoreline and an island), and used Google Earth to measure the distance between the two. Each radar "block" appears to be a square mile. With that information, I made the ficticious radar return below to illustrate my point. I'm not trying to change anyone's opinion, I'm just trying to illustrate that ADS-B returns are not that accurate as-is, especially for smaller returns. In the example below, I've drawn a typical summer return, 6 miles long and 2 miles wide. ADS-B would report it as depicted on the left. A "smoothed" return would look more like the middle drawing (but with straighter straight lines and smoother curves; I'm not a good artist with a mouse), and the return on the right is the actual shower. And that illustrates why I say the current display is not "accurate". ADS-B fills in the entire block when radar detects a certain threshold within that block. Smoothing is also going to keep the majority of the block filled in.
Now let's look at the smoothing. Obviously this is a simplified example, but it's typical. Outside corners will have some of the return rounded off and the inside corners will have a small amount added. In both cases, the smoothed area (either added or removed) accounts for a max of roughly 20% of the width of the block by 20% of the height of the block only in the very corners, and the arc fills in most of that. 20% of a mile is about 1,000 ft. Compare that to your typical takeoff and landing distance - smoothing has removed, or added, less space than most planes we fly can take off or land in. And let's not forget that the return in question is 5-15 minutes old already . At 5 minutes old and 10 mph, the shower has already moved over 8/10ths of a mile (0.85) before the return makes it to your screen. At 30mph, it's moved 2.5 miles in 5 minutes. Adding or subtracting 1,000 ft to the movement (assuming the smoothing is totally inaccurate) becomes pretty inconsequential.
We can argue accuracy all you want, but I really don't think any of us have much of a say in it including iFly. Market movement will dictate the final answer. Just like synthetic vision (which I think is way overhyped and borderline dangerous in the hands of a VFR pilot), apps will be superficially judged and downloaded based on whether they have it or they don't. iFly will have to position themselves to be in the "have it" column to maintain market share. And as OldPilot requested, thats the merits of my opinion.
[attachment]Radar example.JPG[/attachment]