A graph with data quarter by quarter have finer detail.
The steepness of the curve and how is trending during the life cycle already point out if a console had exhausted its sales potential earlier than other consoles.
The problem of comparing long term launch aligned periods by months or quarters is that holydays aren't aligned, so in the case of comparing Switch and PS4 gets weird because they always have alternated bumps. So it's better to do it by fiscal year or to do it by quarters but grouping and aligning them by fiscal year to make sure the Holidays get aligned.
PS4 sell-through in markets like US, Japan and Spain during 2021 is sharply down compared to 2020. In the past fiscal year Sony sold-in 5.80M of PS4 consoles, my guess is that PS4 annual shipment for the current fiscal year will be less than half that amount.
If you look at the long term quarterly Sony sales and compare it launch aligned to previous Sony consoles, you'll notice that in in H2 2020 PS4 had a way bigger drop than previous PS consoles had when their successor was released. And this drop almost matches what PS5 sold.
I think that due to covid and the chip shortage, Sony wasn't able to produce as many consoles as they wanted to match demand of both consoles, but they managed to keep producing the same amount of total consoles they did until then. So they decided to drastically reduce PS4 production and replace it with PS5 production during the launch window to have the record launch and even if they weren't going to match PS4 and PS5 match, at least would have some nice record breaking PR figures for PS5.
To reduce this PS4 stock and shipments slightly before the PS4 launch would also help them to make room on their -and retailers- warehouses and stock for the record PS5 launch that was coming. They announced that were making new factories to increase console production, but that requires time and is still constrained by the chips issue.
I don't know when, but it's fair to assume once they keep solving the chips issue, once they are able to increase production first they will try to balance their PS5 with the demand and after achieved then they'll do the same with PS4 so its sales would be back to normal.
I think they expect that this will happen somewhere in late 2021/early 2022, and this is one of the reasons (plus covid affecting game development, devs always welcoming extra time for polish etc) of why they pushed back GoW or GT7 to 2022, because it's when they expect to have enough consoles on the stores to take advantage of the sale bump these games will mean.
On the other hand you have a console like the Switch which is doing 20M+ consoles sold per fiscal year.
It's already evident that NSW have a high probability to surpass PS4 total sales.
But PS4 isn't really what's Switch is eyeing, what it is aiming for are the PS2 and NDS though, as I've already said, it won't be easy at all.
All consoles have a sales curve, when during its first and last years (where PS4 and PS5 are) they sell less and where somewhere in the middle the console sell the most, has it sales peak. Switch recently had this console peak, which also luckyly for them happened at the same time than the covid bump when people bought more consoles, games and other digital entertainment and unlike Sony they weren't chip constrained.
In fact, if we look at the previous consoles and handheld, it made us think Switch's peak year was going to be 2019 FY because -excluding the extra few months PS4 had- it was mimicking the PS4 sales curve exactly, So even Nintendo expected Switch to sell the same PS4 sold that year. But covid bump plus other consoles struggling for chips happened and Switch sold better than expected and this 2020 fiscal year became its sales peak, instead of the previous one. The sales curve changed and now instead and moved so a mixture of the PS4 sales curve and the DS sales curve.
If things go back to normal the current fiscal year will be the first where after the peak yearly sales curve start to go down first slowly and then faster.
EDIT:
For fun, PlayStation home consoles sales after the release of the successor (taking as a reference the first quarter after the quarter the successor was launched for the very first time, therefore for PS1 it was considered the PS2 japanese launch):
PS1: 29.57M *
PS2: 39.64M *
PS3: ~7M **
PS4: ? (currently 1.00M as March 2021)
Sell-in data unless otherwise specified.
* Production shipment (to not be confused with sell-in)
** Sony announced in early November 2013 that PS3 had reach the 80M milestone. For PS3 not all quarterly data is known due to Sony practice to group PS3 sales with PS2 ones in FY'12.
Yes, Sony home consoles sell around almost one third of their sales after their successor has been released. Because as I mentioned they stay in the market getting supported for longer than the Nintendo and Xbox consoles.
In the past it was more difficult for Sony to support the old consoles because they had to release games made for them and sales were focused on new games. Today instead sales of old games represent a very big chunk of the total (thanks to discounts and so on) and games can be crossgen without being a big pain in the ass, and there's also many services that produces revenue for them from these late gen users without having to invest a lot of money/new games there: PS+, Netflix, Disney+, HBO, Amazon Prime Video, PS Now etc.
During these last years Sony also try to appeal a more family market and emerging markets with price cuts (PS4 is still so fucking expensive), as could be China this time since Sony mentioned of the markets where they want to expand is Asia.
So even if Sony said they will support PS4 during at least 2 or 3 years after PS5 release (that is, to reach the typical 10 years cycle) it may be when they will stop making games or signing 3rd party key games for PS4 they may continue selling it for some extra year more. Catalog games, subscriptions and the typical few remaining yearly games that always are the last ones to support a console like sports of dance stuff will help during these last period.
So knowing that, if chips issues stop soon, if you get PS4 sales and add almost a third you'll have a good estimation of potential total PS4 that would be sold.