Some excerpts from the above article:
Bolt-on armor kits: CDA was asked to develop a set of bolt-on armor for the Sherman, there are pictures of wooden mockups, but this program was canceled before the second gen large hatch hulls started production. At this point, the best source for info on this program is R.P. Hunnicutt’s Sherman. He does not note why it was canceled. It seems like with the success of the M4A3E2 Jumbo, and it’s only marginal effect on the reliability of the automotive components of the Sherman, this would have been a hit with the troops.
Up-armored differential covers: There was another program to improve the armor of the early differential covers. Both the early three-part bolt-together designs and the early one-piece cast designs, were found to have areas more vulnerable to penetration than the rest of the differential cover. They came up with add-on armor for each type. After testing these kits were found to be good enough to make the differentials the best protected front area of the tank after installation. The Army approved them, but no evidence of any being used has been found. The final production cast differential cover was improved and would not have needed these kits. That may have been the reason the kits didn’t get used since they could just use the ultimate production casting when doing rebuilds.
Plastic armor and spikes: When the threat of AT sticks like the panzerfaust become more prominent, an add-on armor kit made from called the HCR2 plastic armor kit was developed. It was made from a mixture of quartz gravel and a mastic compound made from wood flour and asphalt. It was held on by cables and could be jettisoned with ease. The armor from this kit protected the Shermans turret well, but sponson penetrations could still happen. It also offered a little extra ballistic protection. It also did not cover the front of the hull or turret.
Another attempt to defeat shaped charged weapons involved installing spikes in lengths varying from 7 to 8 inches all over the armor. The idea behind this was to break up a heat warhead before it could detonate properly. Testing on this continued after the war.
Fragmentation grenade mounts, mines, and pipe bombs: The Army decided to try mounting these on tanks and test how they would work to combat close in enemy infantry as a kind of last resort weapon. This did not work very well and only the grenades were found to have an effective fragmentation effect. They all risked damage to the tank so they were dropped. Shielding to protect the tank made them even less effective. None of these worked as well as having close infantry support, and the idea was dropped.
The Scorpion/Skink anti-personnel flame projectors: This might have seen use if the war had gone on. This is just the type of thing to use on Japanese suicide troops if they have scared or killed off all your close infantry support. This system had four self-contained, phosphor based, flame projectors mounted at each corner of the tank. Each one could let off 20 to 30 bursts of the flaming phosphorus in a fan from each device, giving great coverage all around the tank. They could be fired off individually or all at once from inside the tank.