Thanks for that. It suggests that the NYC was in fact thinking of a passenger 4-8-4 fairly early on, and not just a more powerful L-4. It had probably noted that D&H had obtained WPB approval for a 75-inch 4-8-4, so perhaps had reasoned that it too could expect to do the same. I suppose that if one started with the D&H K-62, stretched it a bit to accommodate 79-inch drivers, and then enlarged the boiler and firebox to match, one could end up with the Niagara more-or-less. Alco at least would also have had in mind the UP FEF key dimensions as a reference point for a “big” passenger 4-8-4.
Re the Wabash 4-8-2 and 4-8-4 pair, I have never seen an explanation as to why both types were ordered. Worth checking maybe is whether, for example, the 4-8-2 could run over more parts of the system than the 4-8-4. Axle loadings were about the same, but there might have been some underline bridges that could not handle the 4-8-4. I vaguely recall that bridge loading limitation was why B&M opted for a big 4-8-2 and not a 4-8-4.
Or perhaps Wabash ordered its big 4-8-2 before the 4-8-4 was considered to be a suitable freight locomotive. In the pre-4-8-4 era, the 4-8-2 was primarily a passenger locomotive. NYC had also established its [the 4-8-2's] credentials as a freight locomotive, but that idea was not generally taken up. A 4-8-4 freighter would have been really avant garde, perhaps a bold step that the Wabash was not ready to take. So if not exactly stepping out, the Wabash was already taking the path less trodden when choosing the 4-8-2. But then Alco and the Rock Island did step out and introduced a freight 4-8-4, something of a gamechanger perhaps. Maybe the Wabash looked at it, and said “nice idea; we want some of those too”. Its new and up-to-date 4-8-2 would have been the logical starting point for its own 4-8-4.