The record industry has taken such a horrific and unrelenting shoe-ing since 2000 that any glimmer of hope is pounced on and held up as evidence that they did it – that they are through the worst of it. But hyperbolic claims, as made in this Guardian article last week, that it is somehow "saved" are way off the mark. To invert the old Mark Twain quote, reports of its regeneration have been greatly exaggerated.

Yet that was the spin off the back of the IFPI – the international trade body for record companies – recently publishing its global market overview from 2016. (Despite living in a digital age of instant reporting, it still takes several months to pull in all the data from various markets around the world.) The recorded music business saw its global revenues grow 5.9% last year to a trade value of $15.7bn.

The fact of the matter is that the industry had hit the bottom and is only now, like Alan Partridge's sadly pulped memoirs, "bouncing back"... except it still has a long way to go. 2016 was the best performance since 2009 (when the trade value was $15.8bn) but is still a long way off from its prelapsarian time in 1999 before Napster and others kicked it inside out. Back then, the global market was worth $23.8bn.

There is the added fact that in 2001, the IFPI changed what it counted, meaning comparing the reported trade value in 2016 with the reported trade value in 1999 cannot be properly squared. Back then at the start of the great global decline, the IFPI started including performance rights income and synchronisation revenue alongside that from record sales – in part to make the decline seem less pronounced than it was.

Last year, performance rights and synch accounted for $2.6bn – equal to 16.5% of the $15.7bn total. Strip that out and we are left with $13.1bn. When fully comparing like with like in terms of historical data, that "recovery" doesn't sound quite so great now.

If – and, by Christ, it is a big IF – the industry manages to replicate the same level of growth (5.9%) that it did in 2016 (and we'll let them keep their performance rights and sync money in the pot), it is going to take until 2024 until the numbers are better than the $23.8bn reported in 1999. It is only then that anyone can truly claim the record business is "saved". Like a streaming-based Steve Austin, they can rebuild it – but it's just going to take a lot longer than they'd like.

Also, when we look at where the growth was – essentially in streaming as with the myth of the "vinyl revival" is not going to save anyone's arse here – there is a time bomb that is going to fracture growth. For subscribers (so people paying for Spotify, Apple Music, Deezer and – no laughing at the back – Tidal etc.), the best way to explain their value is through the horrible acronym ARPU (average revenue per user) to show what each subscriber is "worth" to the record industry. Last year, there were 112m of them and they collectively brought in $3.48bn – meaning an ARPU of $31.07. Back in 2013, there were just 28m and they collectively brought in $1.12bn – an ARPU of $40.

So, in blunt terms, there are more subscribers but they are individually worth less. This is explained through family packages (getting up to six people on Spotify or Apple Music for £14.99 a month compared to £9.99 for one a single-subscriber package), student discounts and smaller markets where subscription services, due to lower average salaries in these countries, charge way less a month than, say, the UK or the US does.

There are two growth curves happening here – the volume of subscribers and the value of those subscribers. But speed of growth of the former is coming at a cost to the latter. They are not moving in lockstep. So the potential market will become saturated quicker but the average value per user continues to be eroded. All future growth is going to be happening within this context so perhaps the industry should switch from Krug Grande Cuvée to Lidl cava for the foreseeable.

Last year was The Great Harvesting for some of our best pop stars and the industry benefitted from a lot of catalogue sales off the back of celebrity deaths, as always managing to turn a tragedy into a triumph. Prince was the ninth biggest selling artist of the year while David Bowie was the second biggest. Blackstar was part of that, but the bulk of sales were his classic 1970s and 1980s albums as well as greatest hits. On top of this, his triple-disc Cracked Actor album (recorded in, ummm, 1974) was the biggest-selling vinyl set at the recent Record Store Day. So here lies a cruel incongruity for the record business: if it is banking on its biggest stars dying each year to pull it out of a sales hole, the music industry's twist on Logan's Run is only going to go so far until there's nothing left. A thinning crop awaits.

The bigger question is this: who, exactly, is this the early stirrings of a new golden age for? The labels and music publishers? Sure. The top 1% of acts that dominate in CD sales (Adele), streams (Drake) or both (Ed Sheeran)? Of course. But what does it all mean for those way down the pecking order and unlikely to be placed at the top of Spotify's hugely powerful playlists like Hot Hits and New Music Friday?

The three major labels were effusive in their praise for streaming in general and Spotify in particular at the IFPI press launch. No one from an indie label was there but the following day at AIM's Music Connected conference in London, indie label after indie label lined up to heap praise on streaming in general and Spotify in particular. The labels love Spotify. Plus, it should be pointed out, all the majors have an equity stake in the service, as do many indies via Merlin, their collective licensing body. But how is that paying through to the artists? Huge artists love Spotify as it's a volume game. But for those right at the end of the dinner line, it's arguably more famine than feast. And could get worse.

Universal Music and Merlin (on behalf of the indies) have both agreed new licensing with Spotify. As part of that, they can choose to hold back new albums from Spotify's free users for a fortnight. This "windowing" means only subscribers will be hearing albums in their first two weeks of release. As sure as day follows night, Warner and Sony will reach a similar agreement. Isn't that great? Isn't that going to make more people on the free tier want to pay a subscription and thereby increase their ARPU? Well, yes and no.

Reportedly buried in the deal is a clause that, contingent on Spotify hitting certain subscriber targets, labels will take a lower royalty (at the moment, Spotify pays out around 70% of revenues to labels and publishers). So who is going to take the hit here? Will the labels still pay through what they were paying to artists? Or will they reduce their payments to artists in line with their reduced Spotify royalties. No one is saying for definite, but even a cursory glance at historical label/artist deals will tell you who is going to be coming up short here. The bucket might still be dropping into the well but the rope is getting shorter.

For the smallest and most independent artists, brash talk of the industry being saved will surely rankle. If you are on a major or a decent-sized indie, you at least have a fighting chance of making it through the steeplechase to get onto the all-important Spotify and Apple Music-controlled playlists. They are like an accelerated version of the old radio playlist meetings of the recent past where careers could be made or killed before the tea trolley had even left the room. If you are not on them then, in terms of the volume of streams needed to make a living, you might as well not exist.

If you are a small act self-releasing and getting your music on services via a self-serve distributor like TuneCore or similar, it will increasingly feel like a loaded game. Just because your music is ingested onto digital music services, it doesn't follow that anyone will hear it. Standing out from the other 40m+ tracks on there is going to be like the trials of Sisyphus. If Sisyphus was an ant.

In this new digital reality, the big will get bigger and the small will be left hoping for a lucky break, the chances of which get slimmer as each day passes. For them, this is not the salvation being trumpeted by some; this is a three-card Monte.