My answer depends on the time period and information allowed.

 

 

It’s easy, from a post 1945 vantage point, to select the M26 as the solution, or perhaps one of the T-20/22/23 series, because they were do-able with 1942 technology. Unfortunately, this requires knowledge that either was not available at the time or was discounted: the Tiger-I and Panther were soon to be mass-produced, so the M4 needed more armor, and the 75mm wasn’t up to the job.

 

One factor at the time, and perhaps one of the top 2 in importance, was for the US to build as many tanks as possible. Nothing was allowed to get in the way, and production of any of these tanks would have interfered with M4 production. This resulted in about 49,000 M4s being produced.

 

Getting a new and better solution was essentially politically impossible. However, some product improvements could have been made, or introduced earlier.

 

Assuming that the Tiger-I has not yet been encountered in North Africa, the actions are to build /convert all M-4s to diesel and wet ammunition stowage, and probably find a way to add 25mm of armor to the hull and turret front.

 

There might have been enough information at the time to force a switch to the 76mm gun (it fits in the original M4 turret), and retrofit all existing M4s.

 

Changing to the T-23 style turret (better armor) might be possible at this time. The result at this point (1943?) is essentially a fleet of diesel-powered M4A3E8s, with possibly an extra 25mm of armor on the hull and turret front.

 

 

If the Tiger-I has already been encountered, then building the M26 is easier to justify, both at the time and in this scenario.

 

Alternatively, building the M4A3E2 ("Jumbo") with diesel power, stronger power train and a 76mm gun has more merit. The turret has 6" of armor, as cast, the hull is much thicker, and the political repercussions are minimal. Total weight: about 42 tons.

 

 

There is considerable merit to fitting the 17-pdr instead of the 76mm to the M4. Even without APDS, the 17-pdr is very close to the Panther’s 75L70. The 6-pdr was copied as the US 57mm, so doing the same with the 17-pdr should have been a minimal effort.

 

The 17-pdr on an M4 Jumbo looks to be a very interesting combination.

 

Adopting the 90mm, seen in historical hindsight, seems an obvious decision, and it was clear sometime shortly after Normandy that it should have been done. But, in 1942, it’s difficult to see this very-near-term need. It seems that a noticeably larger turret would have been needed, although the M-36 had no problems.

 

Perhaps the 90mm could not fit, perhaps someone with enough influence was opposed to the idea, and nothing could be done to prove them wrong.

 

My own thoughts;

1. I would have gone with the dual purpose 7.5 cm.L38 M-3 gun initially but I sure would have designed a turret that could have taken the later 7.6 cm.L50 M-1.
2. I would have modified the transmission so that I could have reduced the hull height.
3. I would have designed a wider turret ring and a wider tank body[3 inches] so that if I needed that 9.0 cm.L50 I could have it in the upgunning.
4. I would also have gone for wide tracks. The  M-3 Lee/Grant was already criticized for its tendency to bog in the US trials.

Herald
 

I totally agree with those changes plus wet ammo storage in hull bottom (no "ronsons" for me thx), and start work on the HV rounds for the (not being produced yet) 76 mm gun earlier ( not very good anticipation of future armor penetration needs by the US), and hatch modification (drivers hatch changed so turret less likely to block emergency egress, and although the marines put the underside escape hatches to good use picking up wounded on the battlefield, I dont know that that feature contributed to anything other than increased mine vulnerability in europe.

Especially the Sentinel.

Herald

Given the first British battlefield reports, I would, most likely, designed the M4 Sherman exactly as it was in reality.