Skimming through the literature, I find that in theoretical high-energy physics (modern gravitation theories, scalar tensor theories, etc.), older papers are more difficult to read than new papers written in the field, as I spend more time to read an older paper than a fresh one. I wonder why there is such difficulty and contrast for papers written some more than 20 years ago.
I cannot comment on those specific fields, but I had a similar experience with reading old papers in my fields as well with old textbooks on undergraduate mathematics (in particular linear algebra).
The main reason I made out for this is: Didactics marches on:
It usually takes decades for a field to establish a useful notation and terminology. And during that time, new insights may shed new light on what once was considered a reasonable convention – and thus the process repeats. And that’s not even taking into account people sticking persistently to some notation or terminology that everybody else abandoned.
The new terminology and notation usually prevailed for a reason, usually that it makes it more easy to communicate about your field. Also, even if you may be generally familiar with old terminology and notation, you are not as used to reading it, which slows things down.
Similar to the above, scientists have learnt how to write about their fields problems over the decades, allowing them to write more concisely.
Helpful abstractions and generalisations as well as mathematical background are often only discovered after a field has made some progress.
The target readership changed, knowing more about certain aspects of the required background and less about others. For example, many early works on quantum mechanics did not make use of abstract linear algebra as the audience could not be expected to be familiar with it. Nowadays, physicists are taught linear algebra in a more abstract way (due to its importance for quantum mechanics), however, they may not be so familiar with solutions of spherical partial differential equations.