To be honest Wrangler, we could've told you that much! The 1995 to 2005 period was the height of the 'front-loading' marketing ploy that the industry began at the turn of the decade in the UK and perfected within five years, by which time the singles-buying public had become adjusted to the new way in which the industry would promote new singles - i.e. opt for six weeks' radio and TV airplay pre-release, thereby building up whatever interest there was in that song/artist into what they hoped would be sufficient pent-up demand for all interested consumers to rush out and purchase the CD/cassette\download in its first week on sale, and hey presto, maximise its chart potential straight away by (usually) achieving its peak position on week one, often with quite precipitous collapses down the chart from week two onwards, even of those which managed to register a high entry position. This altered the typical behaviour of singles to all-but-eradicate sustained upward movement from any position.
The public were adapted to that form of marketing, but the arrival and eventual integration of the monetised legal digital download began to unravel this, allowing more tracks to grow organically from lower-down the list, sometimes as an album-only cut before receiving a formal single release, and so it started to render the sub-40 and even sub-75 or 100 positions more relevant again in the overall behaviour of songs in the UK charts. Some keenly-anticipated releases from key artists continued to be held back to create the head of steam enough to enter at a higher notch, but since 2015 that has dwindled noticeably, with the industry's focus switching back to the more traditional 'on air: on sale' approach to digital singles, and the rapid rise of audio and video streaming as the dominant means of consuming all music, which has inevitably created a much slower chart with far-fewer entries across the spectrum each week, but with the propensity for very protracted climbs and falls (even with the new rules in place which artificially escalate declines and remove multiple tracks from single artists in order to inflate the positions of fresher titles).
I think to undertake a project on entries below 75 for the 1990s/2000s era would only be something a truly-dedicated chart completist would ever want to do, given the sheer amount of material to compile each week, and they'd need to either be unemployed or retired to have the time to do it!