The difficulty of the exams written by organisations to sell to schools have varied over the years.
At first, they were being sold and used as complete papers, so they needed to be suited to a broad potential candidature, which meant a similar difficulty level as the HSC. That's one reason that the CSSA papers are generally intended to match the HSC standard - they will be used at a great variety of schools, so being much harder or much easier than the HSC doesn't make sense - though some schools using the CSSA papers have changed a couple of questions each year based on topics covered at the Trial. Similar observations can be made of the Independent paper. This also explains why the papers written for high achieving and academically selective schools - NSGHS, Ruse, Grammar, etc - need to be harder than the HSC, as they need to produce a rank order of candidates in a cohort that is generally higher- or high- performing than the HSC cohort.
What I have noticed a lot since the new syllabus, though likely it started earlier than that, is schools purchasing several different papers which allows them to pick and choose questions to create a paper more tailored for the particular school, but still without the needing to take as much time as writing a paper from scratch. Take the 2020 Prairiewood Adv Maths paper, for example - it has questions from the Independent, PEM, and Western Maths papers, as well as some questions that appear to be written in house. North Sydney Boys 2020 Adv Maths has several WME questions, but also one question that is the same as Fort Street (could be a coincidence), a couple of questions in common from the Grafton HS paper, and one in common with Hurlstone Ag HS. This creates more scope for the writers of papers for sale to add in questions at the more difficult end for schools to use or not depending on their cohort. Perhaps the approach seen in the 2021 Two Teach paper of including "Interchangeable Questions" to assist in customising an exam for a particular school will become another standard way to make these purchased papers more useful for different schools will become more common. And, let's not forget the questions that are reproduced from earlier years - questions from CSSA, SGS, and the HSC making an appearance on trial papers a few years later has been happening for a long time.
So, it makes sense to me that the papers for sale to schools will include some more challenging content to broaden the potential market for their product. Schools with stronger cohorts can substitute out some of the easier questions for more challenging alternatives from other papers they purchase. Schools can also choose to use more challenging questions on topics that they wish to emphasise, and to remove questions that they see as inappropriate or undesirable for their individual circumstances - for example, if topic X was rushed or only half done by the trials, don't include 8 marks of question 16 on it.
It is also worth noting that writing exam questions is not nearly as easy as some might suspect. You need to think not only about the solution you expect, but also about alternative approaches and other ways in which a question might be interpreted. There are many threads here at BoS where unintended ambiguities have been discussed, or even questions that cannot be answered without making assumptions about what was intended.
I have no information about PEM (or any of the other companies). They might just have written a more difficult paper this year, or it could be a deliberate change in policy, or it could even be that the perceived difficulty is associated with more difficult questions on a topic where the student doing the paper struggles, or other possibilities. It could also be the change in syllabus resulting in more difficult examples of newer material being explored by exam writers. Then, there are the idiosyncrasies of the exam writers - there have certainly been some weird one-off questions over the years. I can think of a few questions on 2021 science papers that fall into the "I haven't seen something like that before" or the "what was the examiner thinking" baskets.
As a final comment, remember that anyone looking at BoS trials needs to recognise that their purpose is to challenge you and not to reflect the "typical" HSC standard.