Absolutely, this could have been far, far worse, both from an intensity perspective (Cuba) and from a track perspective up the peninsula, with an east coast track being the more catastrophic track, both with regard to the storm surge likely being greater with the stronger side of the storm throwing water onto the land and the much greater population at risk.
I asked a question a few days ago on AmericanWx about almost exactly what you said with regard to accumulated knowledge, although I was asking about that with regard to major hurricanes approaching the US east coast, as my intuition tells me those tend to go east of where they're predicted to go (like Matthew last year). I wanted to know if anyone had analyzed storms of interest, like the much smaller subset of major landfalling/near landfalling storms, instead of just analyzing overall verification scores for any storms. Didn't really get an answer, but often don't in very active threads - may ask it again as a separate thread.
With respect to evacuations and emergency planning, in hindsight, the evacuations of the Keys and Gulf Coast were clearly warranted, while the evacuations of the east coast were mostly unnecessary (although damage there was more than expected yesterday) - I took a bit of a risk and told my dad to sit tight in Vero on Weds/Thurs until we knew the track better (knowing he could high tail it across to Tampa if he needed to in a few hours) and that worked out for him, unlike the many who fled to the Gulf Coast.
But as you well know, with the uncertainty of the science and the need to implement evacuation plans for millions of people 3+ days in advance, what could have been done differently? At 11 am Thursday the model consensus (and NHC track) was for an east coast track and by about 5 pm Friday that consensus was for a west coast track - and for both of those consensus tracks the entire peninsula was still within the "cone of uncertainty."
That's why they essentially felt they had to evacuate the Keys and the east coasts and some of South Florida starting mid-week and didn't really start the evacuations of the west coast, in earnest until Thursday into Friday. Yeah it was terribly inconvenient for those on the east coast, but in the face of uncertainty and potential catastrophe if evacuations weren't ordered, I think it was the right decision.
What made all of this even more complex was that an error of only 50 miles for when the turn to the NW actually happened from near Cuba (and it happened around 8 pm last night), would've made the difference between the track up the west coast and a track up the center of the peninsula, which would've had far more impact on the east coast. That prediction was hard enough to make on Saturday, let alone on Wednesday or Thursday. Hence the conundrum for evacuation planning.
What might be interesting is if they could figure out a way to hold off further on evacuations, but figure out some way to do them much more efficiently (contraflow), so they could be done in 2 days - they maybe then could've held off on evacuations until Friday morning and seen that they didn't really need to evacuate the east coast. But on the flip side, the east coast was still in the cone until very late Friday, which would not give enough time for a fast evacuation. My guess is any storm like this that could come up the peninsula will always have evacuations of far more people than end up being necessary.
Apologies for the long post, but it's something I've thought about a lot and just started typing, lol. The track archive is worth looking at and I posted the Thursday and Friday tracks for comparison.
http://www.nhc.noaa.gov/archive/2017/IRMA_graphics.php?product=5day_cone_with_line_and_wind
Great post-mortem in the link below, although in true
@jellyman fashion, I decided to write up my own post-mortem. It's a little long, lol.
Footage of the damage from Irma in the Keys and SW Florida is sobering. In those locations, Irma certainly lived up to the predictions and most of the people who didn't evacuate are in dire straits without power or water. Through much of the rest of Florida, Irma probably didn't quite live up to the forecasts, as storm surges and winds along most of the rest of the Gulf Coast north of about Naples weren't as bad as feared, as Irma's somewhat unexpected inland track weakend the storm. I'm not going to try to document the damage in this post, as most people are seeing that on TV or the internet.
Having said that, though, Irma ended up having impacts more significant than expected, the day before landfall, for most of central and eastern Florida, as the powerful storm bands on the NE side of the storm led to hurricane force winds/gusts nearly everywhere and fairly significant storm surges on the east coast, especially in northern Florida (Jacksonville, in particular with record flooding) and parts of GA/SC, including Charleston. Of course, the impacts to the FL east coast, GA, and SC were far less than would have occurred if the forecasts from 3-4 days before landfall had been correct with the path up the east coast.
Which brings me to the meteorology of the event, which I've obviously been talking about for days. As all of the experts were saying and I was repeating, there were a few critical points that greatly impacted how this all played out. First, Irma's somewhat surprising landfall and path along Cuba, greatly disrupted Irma's circulation, weakening the storm from about 160 mph to 120 mph. This outcome wasn't forecast that well by the NHC, although he Euro did predict the Cuba interaction pretty well. The Euro was, by far, the best model for the entire storm, as per the graphic below.
Second, probably the most important feature was when Irma finally made its NW, then northward move - this was forecast reasonably well from about Friday onward, as Irma did end up making that turn SE of Key West and generally heading up over the Keys and then up near/along/just inland of the Gulf Coast, as predicted. Obvoiusly, however, this was not that well predicted from 3+ days out, as the forecasts from early last week through about Thursday afternoon consistently showed Irma making that turn, earlier and coming up the east coast of Florida as a Cat 4/5 storm. The difficulty in pinpointing such a turn more than a day or two in advance is inherent to the accuracy of the models, which simply aren't going to be reliable to +/- 50 miles (the width of Florida) from 3+ days out, as was borne out in reality.
Third, the final track from Keys up the Gulf Coast was not forecast that well by the NHC, as the official track generally called for the storm to remain just offshore up to the Tampa area, where the 2nd landfall was expected. As we all saw, the 2nd landfall was made on St. Marco Island, just SE of Naples, and then the storm went up the coast inland 10-20 miles, rather than offshore. This resulted in significant weakening to Cat 1 status by the time it reached the Tampa area, whereas it was predicted to be Cat 2/3 by then and then predicted to make another landfall south of Tallahassee, which never occurred, as Irma stayed inland, heading up to GA/AL.
So, while the forecasts, overall, were pretty damn impressive and certainly everyone in harm's way had more than enough time to evacuate or prepare, clearly, the science of meteorology is still not accurate enough to know precisely where a hurricane will strike several days out (and may never be, due to the uncertainties due to chaos theory, which I've discussed before) and isn't always going to be completely accurate even 12 hours before landfall, as was seen with the surprise inland path on Sunday.
All of which is why people will, for the foreseeable future, need to make evacuation/shelter-in-place decisions with imperfect information, which makes those decisions so challenging. Just imagine another outcome which could have occurred: Irma doesn't come so close to Cuba, remaining a Cat 5 storm, then makes the turn up the east coast of Florida instead of the west coast, giving people only 12-18 hours or so of notice of that deviation vs. the forecast. Talk about a potential tragedy - and that is why emergency response planners will always err on the side of caution in events like this. For a great breakdown of the forecasting of the storm, see the link below.
https://www.washingtonpost.com/news...-room-for-improvement/?utm_term=.060d95ee7f0f