SR N15 vs. SR V schools
both 140km/h
SR N15, 1922 - 1945, $17.82/km. 1800kW x 3.50 = 6300
SR V, 1930 - 1945, $12.54/km. 2000kW x 2.19 = 4400
GWR Bulldog vs. GWR Star vs. GWR Hall
(all 130km/h)
GWR Bulldog, 1899 - 1920, $11.22/km. 1250kW x 3.50 = 4375
GWR Star, 1907 - 1935, $15.18/km. 1600kW x 2.59 = 4144
GWR Hall, 1928 - 1965, $14.20/km. 1250kW x 3.00 = 3750
The Bulldog, despite being an earlier locomotive, outperforms its predecessors by a fair margin.
BR class 20 vs. BR class 24 vs. BR class 30
120km/h to 130km/h
BR class 20, 1957 - 1968, $2.64/km. 745kW x 3.00 = 2235
BR class 24, 1958 - 1961, $9.20/km. 865kW x 4.50 = 3892.5
BR class 30, 1957 - 1965, $4.44/km. 930kW x 3.09 = 2873.7
I would suggest using Gear=3.00 on all locomotives, which would balance out better.
Noted - but I'm going to leave this issue for the time being until I have time for a more serious attempt at balancing speeds, powers and costs. Of course if you want to have a good attempt at this yourself, I'd be more than happy! I think the "gear" parameter is misleading - initially I set these to get reasonable accelerations for typical convoys, but it's hard to tell compare locos easily given simutrans just uses gear*power. I would agree that all locos need the same gear for ready power comparisons.