In my opinion, part of the extreme activity is the number of short-lived storms (less than two days overall - http://www.aoml.noaa.gov/hrd/Landsea/landsea-et-al-jclim2010.pdf - Chris Landsea has a new report on that very topic).
Removing Bret, Gert, Jose, Lee, SS16, Tammy, Vince, and Alpha brings it to 20, which is slightly less extreme. Additionally, it's not likely that Delta, Epsilon, or Zeta would have been tracked or classified pre-Satellite, due to the lack of threat to land, and we would have 17, although that would certainly still be a top ten season pre-satellite.
Less certain would be the intensity. There is an obvious reason that there weren't as many Cat. 5's pre-satellite (and arguably pre-recon) - if they were in the open ocean, the ships would either stay away from the storms, and they very rarely maintain that intensity at landfall. Back through 1943, Recon managed to find some pretty strong storms, including the 1947 Fort Lauderdale hurricane, Dog (1950), Easy (1951), etc. Would they have found Cat 5 status in all four storms? In my opinion, likely not, and I believe Emily would have been missed for how short it maintained that status. That would still leave three.
On a similar level, there would have been fewer major hurricanes. Maria and Beta lasted the shortest, so they would likely not have been recorded properly, Maria for being too far out to sea (and Ophelia being the greater threat) and Beta peaking at nighttime while still offshore. That would still leave five major hurricanes, of which three potentially recorded as Cat. 5's, and none lower than Cat 4. All were major hurricanes long enough that there would be some evidence for their intensity.
As for the record 15 hurricanes, that would also be in jeopardy without today's technology. Cindy wasn't even known at the time to be a hurricane, since there was no recon data, and no land observations of hurricane force winds. Philippe had its peak intensity estimated by Dvorak technique, which wasn't around pre-1974, so it might not have been considered a hurricane (especially since Recon would have been busy with Rita much closer to US). Vince and Epsilon would not have been tracked altogether, leaving a total of 11 hurricanes, still in impressive number. Two iffy ones are Maria (which might have been outside of the area for recon to fly in; additionally, no ships reported a low pressure or strong winds) and Stan (which was a hurricane only six hours before it made landfall, recon could have missed it, and there were no low pressure or strong winds on land). That would leave 9.
A hypothetical 2005 season, pre-satellite (but still in recon era) might have numbers of 17/9/5/3, and I believe that 1933 would still surpass it. Like 1933, however, most of the remaining storms were ones that affected or were near land, or were long-lived enough that they would have been recorded somehow. I don't believe any would have been stronger, with the possible exception of Arlene (if it's 990 pressure was recorded, it might have been upgraded to hurricane status).
This is all extremely hypothetical, and not very consistent, but it's just food for thought while we're waiting for 93L to make up its mind. Feel free to share your thoughts!
