Which console, if any, do you feel missed its ideal launch timing and would've been better off arriving for example a year later or earlier?

In hindsight, PS5 and XBS would have been better delayed a year.

PS5 and Series X/S, too early. Should have waited a year for better stock and games to play...
360 too early, should have fixed that solder problem first
PS3 too early, blu-ray wasn't ready yet, driving up the price
XBox One too late, Kinect was already on the decline and dragged the XBox One down with it. A year earlier and Kinect 2 might have had a chance. (With Kinect 1 supported by XBox One as well, at least for voice control)

Kind of a pattern here, MS and Sony competition leading to bad timing...

These are my initial thoughts. I might pick some others later.
Dreamcast: Delay the launch in all regions by at least a year. Make the specs a little closer to what the PS2 and the others did, and use DVDs. The Dreamcast probably still would've failed, but perhaps not as much.
Xbox 360. Xbox was only 4 years old and faulty manufacturing led to more hardware failure than is acceptable. It also had no HDMI port, which would've been smart futureproofing since most 360s have it. November 2006 would've been fine, especially in the likely event the PS3 still made a bunch of dumb mistakes.
3DS. Delaying it until early 2012 might not have been necessary, but November 2011 would've been great. The 3DS was priced too high, had a sparse early lineup, and no eShop at launch.
Wii U. I'm well aware that the Wii was experiencing droughts in its last years before replacement, namely in 2012. But the Wii U had a very sparse first half of 2013, even for Nintendo games. It would've been better to cling onto the Wii for another year longer and release a few lower to mid-budget Nintendo games on it. Nintendo could then launch the Wii U in November 2013 with a very solid line-up, $300 price tag (no $300 vs $350 options), and an easier to develop for architecture.
PS5: Global supply chain crisis which still would've impacted a November 2021 launch, but would've helped a little.
Xbox Series X/S: See PS5.

Switch 2 if Nintendo releases it in 2024

They need to wait until Summer 2025

 SEGA only had enough money for one option. DVD or the modem and they decided on the modem. I do think DC should have launched worldwide in 1999 rather than in 1998 in Japan and then the US in 99. Tho SEGA as a company was bankrupt. Nothing was going to save them or DC.

I will also add SEGA Saturn launch in 1994 was too early even for internal developers. They were not wanting to move on yet. Saturn has a messy history of it's development from meeting with Sony for letting Sony make it for them to Boeing and more than ending up what the system was and SOA launching the 32X as a bridge gap. Of course, the multiple processes are just too complicated. Whether Sony walked away or SEGA, at least delayed the system to fall 1995 $249 more aimed at 3D with an easy-to-understand architecture and for SEGA to provide developers with proper tools. (they were notoriously bad at this) from other internal studios to 3rd parties)

Turbo Grafx 16/PCE. Before Mega Drive and right when NES/Famicom is ramping up. it had no chance.

Apple pippin

GameCube is pretty much the test case for this.

It really, really (reeeeeeeally) needed to launch in like November 2000, which would've been a year earlier. 

Giving the PS2 a full year+ headstart (18 months in Japan) was too much and then having it release alongside the XBox while XBox had a killer app in Halo and the GameCube really didn't have a game on that level was a big problem.

GameCube should've launch in US/Japan in November 2000 with Zelda: Majora's Mask and Perfect Dark (taken from the N64, lets face it, Perfect Dark barely ran properly on the N64) as launch titles.

I think if this happens, do they beat Sony? No. But they probably would've had a healthy no.2 position, and XBox would've kinda been DOA, far too late in the game as GameCube would have a larger userbase with the year head start. 

The virtual boy was released about 20 years too early

Ideally, you have to release a console the year there's a die shrink available so you maximize performance for cost and longevity (as well as having a noticeable improvement over any competitor that might have released earlier). The PS5 and the XBX would have benefited nothing from delaying a year, it should have been two or nil in this case.

The PS3, on the other hand, would have gained hugely from waiting 9-12 months and going to the 65 nm node and unified GPU architecture. Also much cheaper Blu-ray drives, too. The opposite would apply to the GameCube, since it could have been released earlier with the same specs and better compete with the PS2.