Use of RP25 wheels in US H0 plastic steam RTR locos

If it's all current (new or retooled) production, yes. There may however be subtle differences in interpretation of the RP25 specs - back-to-back and check gauge dimensions are more important the flange depth for good tracking, especially through turnouts, for example.
OTOH, on some older Rivarossi and IHC locos the pilot and trailing truck wheels are not RP25.
Well, retooling is costly, so why do it if the tools are still good, and so many people want cheap trains? Also, "common sense" says that deeper flanges will hold the track better, right? Exhaustive NMRA testing in the 1960s disproved this, but hey, what's accurate testing data when it disagrees with to common sense? ;-)
Avoid any "entry level" locos, whether steam or diesel. They are made to a price point, for people who have a bad sense what inflation had done to prices. There's a reason so many of these locos are offered at 50% or more below "suggested retail price". Better buy one good loco than three "blow out sale" junkers.
Anyhow, the RP25 issue is a mix of aesthetic, technical, and economic considerations. RP25 is not an industry standard, it's a recommended practice of the NMRA (National Model Railroad Association, a volunteer _consumer_ group.) The NMRA cannot enforce its standards, only market forces can do that. It also helps to join the NMRA, so that it can continue to develop, improve, and promote standards.
The comments below are in no particular order.
-- RP25 was developed by the NMRA primarily for aesthetic reasons -- the shallower flange looks better. The NMRA flange standard specifies a maximum depth of 0.035" and a minimum of 0.025". The deeper flange will track properly on code 83 and larger rail, and even on some code 70 track. For example, Micro Engineering's track has finer "spike" heads than some other brands, and code 35 wheels will run just fine on it. (Code numbers refer to thousandths of an inch.)
-- The NMRA standard wheel profile specifies a fillet between flange and tread, for all flange depths, and a fairly thick flange. (There is also a "fine scale" spec for a closer to true scale wheel.) This rounded area helps prevent the wheel from climbing the rail and derailing. Unfortunately, "RP25" has been interpreted by some manufacturers as referring only to flange depth.
EG, I have some "RP25" wheels sets which have the shallower flange, but it is narrow and sharp, and there is no fillet between flange and tread. I keep them as curios, to show what happens when a manufacturer doesn't fully understand what the NMRA standards and recommended practices mean.
-- NEM allows a deeper flanges (max. of about 0.045"), which explains the generally very deep flanges on European made locomotives in the past. But recent European and British product has better looking wheels, close to or following the RP25 specs. AFAIK, the Japanese manufacturers adopted RP25 almost as soon as it was approved by the NMRA membership, largely because so much of their market was the USA/Canada at the time.
-- Market forces do eventually have an effect. All new quality locomotives have RP25 wheels, just as all new locomotives now come at least "DCC ready", and the vast majority are offered in both DC and DCC versions which actually raises the average price - it would be cheaper overall to offer all locos with DCC installed.) Also, more and more of the older product is wholly or partially retooled, and that means RP25 wheels
-- "Entry level" now means about $100 for a locomotive, and that's the discounted price. There are exceptions, eg IHC, but I consider those as trial gear - good enough to last long enough for you to decide whether model railroading is for you, cheap enough to toss when you decide either way. $100 to $150 looks high, until you consider that entry level locos in the 1950s ran around $25 to $30, or roughly $200 in today's money. And those were kits. I recall the buzz when Varney offered his Dockside 0-4-0T at $15.95 - still over $100 in today's money.
HTH

Thanks.
Here (in Finland) the "average" modeller runs with german commercial trains and belives in MOROP's NEM. All finnish outline is considered by average enthusiast as "finescale" and rivet counting although all kits produced have only RP25/110 or RP25/88+ wheels (as practically nothing is available in RTR in Finnish outline and those wheels are most widely available). Small group of us have tried to convince that even kids in US drive on layouts laid on carpet with RP25 and are happy.
I therefore wanted to check if this really is true, and that the kids with their first train set having a steam loco with more that two axles can actually keep the RP25 wheels on metals.
It appears that it is not so simple as old tools and models are around (models are often made in big batches for years to sell), but that there seems to be no technical reason why RP25 would not keep on track even laid around the xmas tree.
It seems that the difference between MOROP's NEM and NMRA's standards is that in Europe NEM tries to follow the manufacturers factory standards but in US manufacturers try to follow NMRA standards...
pekka

(snip) From helping friends with collections of old Marklin and Fleischmann trains, I have the impression that both systems first (or at least early post-WWII) used track that was made with sheet metal rails bent into an inverted "U" shape that was tab and slot mounted to a sheet metal base with a "roadbed" trapezoidal cross section. This semicircular top required a deep flange to keep the train on the rails, just as with US Lionel and its tubular rail head sheet metal track.
By contrast, I believe that HO started in the US as a craftsman's hobby, and from the beginning used solid rail with a sharper angle at the top corners of the rail head. Most all of the early track (MidLin, TruScale, Atlas flex, etc.) were designed for use on permanent layouts and were not originally offered as sectional track systems. I think these factors resulted in the adoption of a more prototypical flange in the US. I recall that it was not until the mid-1950's that Atlas introduced the "Snap-Track" sectional track system.
(Does anyone know the commercial history of this event - did Snap-Track open up the availability of HO train sets, or did the HO train manufacturers' desire to offer train sets lead to the introduction of Snap-Track?) Geezer

[...]
OK, see your point.
Yes, there are cheap entry level train sets here with the larger flanges (0.035"), but otherwise the wheel profile is usually standard NMRA. The better train sets use RP25 wheels.
Most people don't run on the carpet, though. There is track mounted on plastic bases that vaguely resemble ballast, and that's OK for carpet running, but even so, the first layout is usually built on a 4'x8' (120cmx240cm) sheet of plywood.
The NMRA was formed originally (in 1936) to promote interchange. There were many different scale modelling standards, and modellers found that they could not be certain that their lovingly built models would run on other people's layouts. NMRA standards fixed that problem. This had nothing to do with toy trains, whose only standards were track gauge (1-1/4" or 32mm) and propulsion voltage (16v AC, more or less.) Scale models were in a minority, served by small manufacturers, many of whom were themselves model railroaders. These welcomed common standards, which guaranteed the largest possible market for their products. Eventually, the toy train market withered, and the scale model market increased.
So, from the beginning, the NMRA focussed on promoting standards for manufacturers to follow. MOROP was formed after the fact, so to speak, and so had to adapt its standards to the de-facto ones used by the manufacturers. But I've noticed that European products are increasingly using RP25. I think the main reason is the increasing demand for accurately made models, which higher disposable incomes have made affordable. Model railroads have become a largely adult hobby (which does not bode well for its future).
I suspect that Maerklin's recent problems were as much caused by its image as a toy maker as by a backlash against its high prices for what became at best average product compared to other brands. Maerklin still insists on going its own way, which will cause it further problems. The fact is that an international model railroad standard, both technical and aesthetic, is coalescing, and manufacturers that ignore it do so at their peril. The days of captive, national markets are over.
Anyhow, the history of model railroading has never been fully documented or written.
Have fun!

Well, actually NMRA found that a wheel with no fillet between flange and tread was required for good tracking on round-head rail, and that shallower flanges would in fact work on such rail. Conversely, a filleted wheel tracked well on square edge rail. That promoted both a revision of the NMRA wheel profile, and allowed a reduction in flange depth. More recently, several workers have shown that exact scale wheels will work perfectly, if trucks and chassis have working equalisation, and track is built to matching standards.
Good points.
IMO Atlas's Snap Track (early 1950s) was a response to foreign imports. IIRC it was the Italians who first offered solid rail sectional track with a realistic rail profile. Around that time Fleischmann also offered solid rail sectional track. IOW, the furriners showed it was possible, and Atlas copied and improved on their innovation. For decades, Atlas was the standard, and other manufacturers had to make compatible track. Hence the code 100 sectional track on plastic ties that has become a de-facto standard worldwide. We are now seeing a competition between varieties of plastic ballast base track. And code 83 rail is slowly displacing code 100.
Fun times. ;-)