Occasionally, when preserved steam locomotives are run around people who know diesels, someone compares them. I remember two comparisons that got into print.
(i) EMD put some money into the American Freedom Train project in 1976, and so had people involved. A report in (I think) "Trains" quoted an EMD person as saying that the (ex-SP 4-8-4) used was, when it came to running the train at a good speed over various grades, equivalent to two and a half E-8 (so: to diesel locomotives rated 5625hp). Caveat: this was at speed. When it comes to starting the train from a dead stop, big 4-8-4 typically had a tractive effort of around 60,000 pounds. Two and a half E-8 would pull about twice that.
(ii) An ex-NKP Berkshire was run on a CSX line; CSX assigned a B36-7 to help. I don't remember the details, but apparently there were times on a run when each locomotive did the work. A story in "Railroad Model Craftsman" said that ONE B36-7 wasn't equal to the Berkshire.
---
When the SD70-MAC was introduced, one historically-minded EMD person said that they "finally" had a single unit that could out-pull a (Norfolk & Western) Y6b (the final best 2-8-8-2 mallet): I think this was a tractive effort comparison, not horsepower.
---
Early in dieselization, the New York Central did a carefully planned comparison of its "Niagara" 4-8-4 with EMD E-7 on Croton (New York) to Chicago passenger trains, famously finding the costs very similar (but maybe stacking the deck a bit by choosing test runs EXACTLY taylored to the steamer's characteristics). The Niagara significantly outperformed two-unit E-7, and was similar in performance to a three-unit E-7. The Niagara's horsepower (New York Central tested its locomotives very carefully) was over 6,000 "indicated" (basically power produced at the cylinders) at its best speed: power at the back of the tender was (i.i.r.c.) either 5300 or 5100 at best speed. (Steam locomotives only produce peak horsepower in a comparatively narrow range of speeds: one of the selling points of the diesel was that it could use full power over a broader range.) A three-unit E-7 is rated at 6,000 hp, under the (American) convention that the power of a diesel locomotive is the power the engine delivers to the main generator: with first and second genereation diesels, the train-pulling power was about 83% of the rated power: 4980.
---
I vaguely recall seeing an estimate of about 500hp for a mid-19th century 4-4-0. Trains were much lighter then!