First off, forgive any typos, I'm on mobile atm!
I'm a big fan of narrow formations in general and the 442 diamond/4312 (my preferred nomenclature) holds a special place in my heart. It's a setup that seems out of fashion in modern football, likely owing to the inherent issues defending the flanks and the rise of Guardiola inspired "positional play" and it's emphasis on on maximum width to stretch opposing defenses. Despite that it can be a potent weapon in the hands of a manager that understands the setup and a team capable of meeting it's specific demands.
Let's start with the strengths of the formation. With four central midfielders the 4312 will always have at least parity and most often superiority in the center of the pitch. When defending this can force the opponent to progress the ball on the flanks, a more circuitous route to goal. When a team utilizing the 4312 has the ball, there will almost always be a free player in the midfield to receive a pass. This superiority in the middle of the park is not merely an advantage but the 4312's entire raison d'etre.
When in possession, the 4312 allows a manager to field both a two player strike partnership pressuring the backline and a creative #10 between the opposition's midfield and defense. This attacking trident can be very difficult to effectively shut down, particularly for teams playing a back four.
In attacking transitions, that trident allows the team playing a 4312 the option of playing the ball forwards very quickly. One of the center forwards can drop deep to receive the ball with back to goal, the other pinning the center backs, and the #10 lurking in dangerous areas in anticipation of a flick on or wall pass.
It's worth taking a moment to point out that, with a holding midfielder and two flankers, the #10 in a 4312 is often relieved of the defensive responsibility they would have in more modern setups, i.e. 4231 or 433. While one can certainly play a hard working, team oriented player in that role, it is not as necessary as in other setups, allowing for the use of a "luxury" creator in that role.
On to the weaknesses. First on that count is the fragility on the flanks when defending, particularly against the opponent's attacking transitions. If one's team loses the ball in the opposition area with both fullbacks committed forward it can be all too easy for an opponent to play quick balls behind the fullbacks for their wingers to run onto. In instances where the opponent has controlled position in one's own half they can look to exploit the narrowness of the formation by utilizing wide attackers and overlapping fullbacks to create a 2v1 with one's fullbacks. Whether in transition or against settled posession, how a team utilizing the 4312 handles pressure on the flanks is often the difference between victory or defeat.
The other weakness is much like the first, but flipped. Against teams that field an organized mid/low block, width in attack can become a problem. Even with two fullbacks bombing forward it is undeniably easier to pull a set defense apart with numerical parity.
Finally, and this is more a team building issue as opposed to a tactical one, those fullbacks had better be up to the task, particularly if one's side will look to dominate the ball and play in the opponent's half. They have, bar none, the most physically demanding role in the side. They can, and will, tire quickly. Thus it is imperative to have at least two options for each flank, ideally with the drop off from first choice to second kept to a minimum. When playing a 4312 I find myself subbing off one of the fullbacks every game, and often both. Fixture congestion can can exacerbate the issue.
With that out of the way, there is a lot of wiggle room as far as how one sets up their own 4312. Concerning the example you've given, it looks to me like you broadly want to play a somewhat direct counterattacking style with a mid block. In light of that I'd make the following observations...
1. On a positive mentality, both of those WB(s) are going to be quite aggressive. They're going to push high up the pitch early in posession when compared to a FB and when they get the ball look to drive at the opposition at pace. Nothing inherently wrong with that but I would consider either a FB(a)/FB(s) combo or perhaps WB(s/a, depending on the quality of player going forward)/WB(d). Don't sleep on the WB(d), they will generally push higher earlier than a FB(s) while being less likely to bust down the flanks, instead working more as a safe option for ball circulation that can drop in crosses from deeper positions. I've had good success using one of them in the 4312.
2. You mention getting undone late in games. That strikes me as a fitness issue, particularly with the wingbacks. Watch their fitness and sub them off when they're exhausted.
3. A vulnerability to through balls is, in my opinion, a big red flag. You shouldn't be having this problem in a formation with four midfielders. While I've not used the role in FM23, I found that in previous versions the HB was problematic, particularly in narrow setups. The role will drop between the CBs in buildup, splitting them and encouraging the FBs forward. That's all well and good, but in sustained posession the HB would step into midfield and the CBs would stay split. Not sure if that's the issue you're running into but I would suggest looking at those goals and trying to suss out if that's the problem. Personally I've used an Anchor, DM(d), DLP(d) or CM(d) to good effect. I'd avoid BWM, they're a bit too prone to ball chasing IMO. The other thing I'm thinking is that, with the close down more instruction, your midfielders are getting dragged out of position. Can't tell without seeing the goals conceded.
4. I don't see any roles that create for others. GK, CD, HB, CAR(s), AM(a), PF(a), AF... none of those are inherently creative roles. I would encourage you to look for players in the team who can move the ball and unlock defenses and use them in roles that encourage them to do so. If you look to play one attacking fullback and a more conservative one don't be afraid of a MEZ(s, maybe even a) on the side of your more reserved fullback.
5. Speaking of creating, that attacking trident is three roles that are going to primarily attack the box. There's nobody tasked with dropping in and holding the ball, running at the opposition, or playing killer balls. It's going to depend a lot on the players available to you but there should be some diversity there. Something like AP(a), DLF(s), AF can be very effective. One forward pinning the defense, another showing for the ball and playing others in, plus an attack minded playmaker just behind looking to run at opponents and find that pass to unlock them.
For reference, the most effective 4312 I've used was in FM22 and it looked like this.
GK - SK(d)
RB - WB(d)
CB - CD(d)
CB - BPD(d)
LB - WB(a)
DM - A
RCM - MEZ(a) *PI - Run wide with ball, cross more often*
LCM - CAR
AMC - AP(s)
STRR - AF
STRL - DLF(s)
I don't have the tactic in front of me but it was broadly posession based with some counterpressing. Positive mentality, play out of defense, shorter passing, counter press, counter, higher LOE/DL. Nothing too crazy but it was effective at pinning most sides deep and dominating play, though better UCL opposition was a challenge. I would note that I did NOT use WBIB, I don't see the point of that instruction with two forwards and the counter instruction, particularly with a #10 behind them.
For playing on the counter, I'd suggest keeping the same TIs as you have, add Counter, drop press more often if it's causing the through ball issue, add get stuck in, maybe hit early crosses, using the roles above but dropping the MEZ to s, LWB to s and bumping the AP to A. I'd also consider asking the GK to go direct to the DLF, perhaps even using a TF if that better suits your squad.
Good luck, and I hope the 4312 serves you well in the future!
So @nick1408, I definitely ran into similar problems against high level UCL teams in my FM22 save while using the 4312. 1. You posted scores from three games that have your side at 12 goals scored to only 2 conceded. That's a hell of a return. Context is important and the sample size is small but I would be quite happy with those results. 2. As to defending the fullbacks from those goal kicks... I'm not sure! As I mentioned my squad had some issues against top level UCL competition. I wound up living with it as my board didn't require more than a respectable showing in the UCL and I managed that rather comfortably despite not advancing out of the group stage against PSG, Dortmund and Ajax. This is just theorycrafting, but I would consider either asking at least one of the strikers to mark the fullback on their side, or bumping the A(d) to a CM(d). The latter will have the effect of making your CMs wider and potentially get quicker pressure on the FBs, the former would tend to force the opposition to building up on a given side of the pitch where, hopefully, your midfield can shift over and snuff out their attempts to build from the midfield into the attacking third. In my experience, the weakness defending the flanks was worth it for our goalscoring output. With Sporting CP I managed around 3.5 goals/game in the Primera Liga, more than enough to dispatch most opposition with ease. Even against better opposition we still scored, but those sides in the UCL definitely had an easier time holding on to the ball and creating chances. Your mileage may vary, but IMO it's worth sticking with unless you're under serious pressure from the board to shore up your defense. The quantity and quality of goals my squad scored was intoxicating!