If some of the television meteorologists did not properly explain the threat, I think that's more on them. The NHC does not handle flood watches and warnings. There's even other wind watches and warnings handled that are not usually discussed in the NHC advisories. Tropical storm watches and warnings have nothing to do with rainfall or surge, only 1 minute sustained wind of 39mph or greater. (though that does now include: tropical, subtropical, or post-tropical cyclones)

The cone is simply a metric that says over the past 5 years the center point of storms have stayed within it two-thirds of the time. One-third of the time the storms did not. It does not take into account the size of the storm. It doesn't mean that two-thirds of the time impacts will not be felt outside the cone. The storm could be within the cone, on the very left edge, with impacts hundreds of miles beyond that left edge of the cone.

Even in this particular meteorologist's post he still doesn't explain this well:

"The storm could go anywhere in the cone, and even, on some occasions, outside the cone. In fact, the current day-five margin of error in a tropical cyclone's position is still over 230 miles."

He should have explained it to people better. Saying "even, on some occasions, outside the cone" is not an explanation. His post still has not explained it and this was the perfect opportunity to. One-third of the time does not equate to "on some occasions" to me. And then there are the impacts outside the cone, because a storm is not a point.

The NHC did forecast the wind field to expand and it did. The NHC was a little late to mention Virginia in the rainfall threat perhaps, though I don't know how local forecast offices communicated the threat. I remember seeing flood watches beyond the Carolinas though I forget when and they may have been in effect well before. It was well known in advance that this would be a serious rainfall threat. The NHC often said the Carolinas though, so in that regard they could have mentioned Virginia earlier. Matthew was off the coast of Georgia before they started mentioning up to 15 inches in Virginia.

This was not simply about Matthew, it was about the front too. It was always clear that would also contribute to the rainfall, though the NHC did not anticipate extratropical transition much in advance and that did aid in more the rainfall.

I don't think there should have been tropical storm warnings issued in Virginia. You could make an argument for a watch since it includes post-tropical cyclones, but it was handled by the NWS in other ways.

In fact the NHC even said this at one point: "Strong winds in the Tidewater Region of Virginia are being handled by non-tropical wind warnings."

Meteorologists on television are the ones that usually interpret that and then communicate that to their audience. The NHC can't cover everything, that is why local NWS forecast offices are essential. I would need to have seen when specific watches and warnings came out from those offices to see how well they did, but it sounds like this meteorologist also got caught up by the cone. Local NWS forecast offices need to be the ones to communicate local threats. They know their area. That is why in a storm like people need to look directly at what local NWS forecast offices are saying, if the meteorologists on TV are not doing a good job and that is usually how you get information.

It's just important that people rely on local NWS offices for such things as flooding and other impacts locally. The NHC can't get into all that extra detail for every locality.

A lot of local planning is done by coordinating with local NWS forecast offices for local area impacts. You're right about people rightfully or wrongfully going crazy regarding the cone, but that can't influence decision making. There should not be tropical storm watches and warnings just so people pay attention. It should be used when the criteria is met. Maybe more people would have paid attention in Virginia with actual tropical storm watches and warnings, but unless actual tropical storm conditions are possible (watch) or expected (warning), I don't think they should be issued. Then that makes people not trust those issuing them.

I think issuing tropical storm watches and warnings for post-tropical cyclones can really confuse things. There are other wind watches and warnings that can be issued. A nor'easter would be handled by those kind of warnings and not tropical storm watches and warnings, yet you could see a similar type of storm in the end, that originated in the tropics, do get tropical storm watches and warnings. That change is due to Sandy and people not paying attention to local products issued. Based upon the pressure, the NWS made changes:
http://www.nhc.noaa.gov/news/20130404_hsu_postTropicalChanges.php

I don't know, maybe people expect tropical storm and hurricane watches and warnings for every storm originating in the tropics now, or maybe that is what they have always assumed. There used to be clearer differences and not as subjective when to issue a tropical storm warning or some other wind advisory.

Watches and Warnings defined:
http://www.weather.gov/lwx/WarningsDefined

As for south of Haiti, the NHC is not great at intensity sometimes. They definitely admit that intensity forecasts are still lacking, while track forecasts have gotten better over the years. They did get that very wrong here, though the NHC was better with the track. The NHC was good with the track about two and a half days out from Haiti. The forecast movement after the Carolinas was an exception. They definitely highlighted uncertainty though regarding the potential for the storm to begin moving southward after the Carolinas.