This website offers some statistics. The latest set they have covers 2025 so doesn't include the Berlin/Paris service but it does cover a year's worth of the Brussels/Prague one. You can only filter the data in certain ways, but perhaps a sort-of meaningful way to do a comparison is to look at trains that pass through Germany. After all, we know that Germany is currently a major source of delays, so we can look at how a number of operators do when trying to run long distance trains through the country.
You can order the results by Average delay, cancellation rate or punctuality. I think the first two are the most meaningful for night trains - as you say yourself, an expectation of the possibility of delays is part of night train travel.
Ordered by "Average delay" the worst service they list is EN 345 Stockholm-Berlin with an 84 minute average delay. It's followed by a couple of Nightjet services at around 60 minutes. You can choose to display 50 results per page. The top 50 results are mostly a mixture of NJ, ICE and IC and EC services. The average delays range from 84 to 37 minutes. Are the ES services in that top 50? No - in fact you have to go to page 6 of the 26 pages of results to find ES 452 Prague - Brussels with an average delay of 25 minutes and page 11 to find ES 453 Brussels - Prague with an everage delay of 20 minutes.
You might say that the most critical statistic is "cancellation rate" because of course cancellation of a night train usually causes you a much bigger problem than a delay does. You can order by cancellation rate too, and the worst performer is an ICE with 35.6%. ES 452 Prague-Brussels is on page 20 of 26 with a cancellation rate of 2.4% and ES 453 is on page 24 of 26 at 1.1%.
For the worse of those two, ES 452's cancellation rate means that if you travelled on it once every week, you could expect your train to be cancelled once a year.
You may say that comparing cancellation rates with day trains is a bit meaningless because of the magnitude of the consequences and I'd agree, so it is fairer to compare with other night trains. Here are all the night trains that are listed as passing through Germany:
| | | | | |
| Train no. | From | To | Average
delay (mins) | Cancellation
rate (%) | Punctuality
(%) |
ES 452 | Praha hl.n. | Bruxelles Midi | 25.6 | 2.4 | 52.0 |
ES 453 | Bruxelles Midi | Praha hl.n. | 20.6 | 1.1 | 63.4 |
| | | | | |
EN 346 | Berlin Hbf | Stockholm Central | 12.8 | 5.2 | 81.1 |
EN 345 | Stockholm Central | Berlin Hbf | 84.4 | 4.4 | 29.1 |
| EN 13485 | Stockholm Central | Berlin | 77.8 | 5.7 | 29.8 |
| | | | | |
EN 40459 | Zürich HB | Praha hl.n. | 37.8 | 1.7 | 48.5 |
EN 40458 | Praha hl.n. | Zürich HB | 36.3 | 0.9 | 55.2 |
EN 40476 | Budapest-Nyugati | Berlin Hbf | 19.2 | 1.8 | 58.3 |
EN 407 | Warszawa Wschodnia | München Hbf | 18.4 | 1.0 | 37.8 |
EN 414 | Zagreb Glavni kolodvor | Stuttgart Hbf | 16.2 | 0.4 | 68.5 |
EN 40457 | Berlin Hbf | Budapest-Nyugati | 14.7 | 1.8 | 70.7 |
EN 40237 | Stuttgart Hbf | Zagreb Glavni kolodvor | 11.8 | 3.1 | 78.4 |
EN 406 | München Hbf | Warszawa Wschodnia | 6.9 | 1.2 | 79.9 |
| | | | | |
EN 50462 | Budapest-Keleti | Stuttgart Hbf | 16.2 | 0.4 | 67.1 |
EN 50237 | Stuttgart Hbf | Budapest-Keleti | 11.4 | 2.1 | 79.8 |
| | | | | |
NJ 40424 | Berlin Ostbahnhof | Paris Est | 26.1 | 9.4 | 55.4 |
NJ 479 | Bruxelles Midi | Mannheim Hbf | 30.2 | 7.4 | 38.6 |
NJ 424 | Berlin Ostbahnhof | Bruxelles Midi | 31.6 | 6.7 | 55.0 |
NJ 40425 | Bruxelles Midi | Wien Hbf | 42.5 | 6.0 | 40.8 |
NJ 40469 | Paris Est | Berlin Hbf | 27.4 | 5.9 | 55.2 |
NJ 425 | Bruxelles Midi | Berlin Hbf | 41.8 | 5.9 | 44.8 |
NJ 468 | Wien Hbf | Paris Est | 28.6 | 5.2 | 36.3 |
NJ 40468 | Wien Hbf | Bruxelles Midi | 38.8 | 5.0 | 47.7 |
NJ 402 | Zürich HB | Amsterdam Centraal | 23.8 | 4.8 | 56.9 |
NJ 478 | Mannheim Hbf | Bruxelles Midi | 26.4 | 4.5 | 41.9 |
NJ 403 | Amsterdam Centraal | Zürich HB | 42.8 | 3.6 | 33.8 |
NJ 421 | Amsterdam Centraal | Innsbruck Hbf | 51.7 | 3.0 | 38.4 |
NJ 40421 | Amsterdam Centraal | Wien Hbf | 54.8 | 2.9 | 31.1 |
NJ 40295 | München Hbf | Milano Lambrate | 7.3 | 2.6 | 76.3 |
NJ 40490 | Wien Hbf | Amsterdam Centraal | 62.9 | 2.5 | 20.3 |
NJ 295 | München Hbf | Roma Tiburtina | 8.8 | 2.2 | 75.9 |
NJ 40491 | Hamburg-Altona | Innsbruck Hbf | 45.8 | 2.1 | 60.7 |
NJ 237 | Stuttgart Hbf | Venezia Santa Lucia | 12.3 | 2.1 | 75.6 |
NJ 470 | Zürich HB | Hamburg-Altona | 28.9 | 2.1 | 43.9 |
NJ 471 | Hamburg-Altona | Zürich HB | 26.6 | 2.0 | 63.4 |
NJ 469 | Paris Est | Wien Hbf | 36.4 | 1.9 | 33.3 |
NJ 408 | Berlin Hbf | Zürich HB | 32.5 | 1.7 | 61.3 |
NJ 420 | Innsbruck Hbf | Amsterdam Centraal | 64.0 | 1.7 | 25.0 |
NJ 409 | Zürich HB | Berlin Hbf | 39.3 | 1.6 | 46.8 |
NJ 491 | Hamburg-Altona | Wien Hbf | 54.2 | 1.4 | 55.7 |
NJ 490 | Wien Hbf | Hamburg-Altona | 51.6 | 0.9 | 46.8 |
NJ 40420 | Innsbruck Hbf | Hamburg-Altona | 52.5 | 0.9 | 47.7 |
NJ 456 | Graz Hbf | Berlin Hbf | 20.5 | 0.8 | 56.7 |
NJ 236 | Venezia Santa Lucia | Stuttgart Hbf | 32.4 | 0.8 | 34.1 |
NJ 457 | Berlin Hbf | Graz Hbf | 15.1 | 0.8 | 70.1 |
NJ 294 | Roma Tiburtina | München Hbf | 24.5 | 0.7 | 56.1 |
NJ 40235 | Milano Lambrate | München Hbf | 28.5 | 0.0 | 55.9 |
I don't think you can look at these figures and say that compared to other night train services, European Sleeper is some kind of major outlier when it comes either to delays or cancellations. In fact I'd say they compare pretty favourably and appear to be better than average.
I don't know anything about how these data are collected or how reliable they are. And there may well be ways in which things are reported that distorts reality in some ways but this appears to be the best there is in terms of objective numbers. I think I'd pay much more attention to this than to a general feeling from social media or youtube trip reports.
A couple of years ago I did something similar on the Caledonian Sleeper thread, comparing the sleeper to Inverness with flights, looking at delay and cancellation rates. Because so much more attention is given to problems with the sleeper than the airlines, it would be easy to get the impression the sleeper was terribly unreliable. But the real world statistics didn't seem to bear this out at all - it was no worse than the flights and if I remember correctly actually a bit better. And it looks to me like the same is probably true here - because of the nature of the operation ES gets a lot of attention and what gets published (including in reviews) is heavily biased towards the times when things go wrong.
Perhaps next year we'll be able to look at a similar set of numbers that includes the Paris/Berlin ES services.