[Bug]: Confirm/Alert popups break focus

[LqdBcnAtWork]

Preflight Checklist

• I have read the Contributing Guidelines for this project.
• I agree to follow the Code of Conduct that this project adheres to.
• I have searched the issue tracker for a bug report that matches the one I want to file, without success.

Electron Version

29.1.4

What operating system are you using?

Windows

Operating System Version

Edition Windows 10 Home | Version 22H2OS build 19045.4170 Experience Windows Feature Experience Pack 1000.19054.1000.0

What arch are you using?

x64

Last Known Working Electron version

No response

Expected Behavior

Using a popup (alert or confirm) then trying to focus an input element shouldn't cause issues.

Actual Behavior

When the application is launched, there are no problems interacting with an input element.

After clicking on a popup from confirm or alert, trying to focus the window that just caused the popup (namely an input element) will no longer work.

Focusing on another window and refocusing the window repairs the issue.

Testcase Gist URL

https://gist.github.com/LqdBcnAtWork/9af23d4faa33e74874cd08ffbc8e4ed6

Additional Information

I've noticed this issue for a while, but didn't catch onto the issue being caused by popups. I thought it was random and was caused by some strange interaction with focus that I was causing on my end. Having diagnosed the problem's cause, creating a Fiddle that could replicate it took seconds.

I mention this because it means the issue has existed for many major versions of Electron. How far back is unknown to me. But it goes at least as far back as version 21. I tried version 12.2.3 for fun with Fiddle and the problem persists.

I believe this is caused by the popup taking focus, but the window doesn't recognise that it gets focus back when the popup is closed. I also think this issue might be on Chromium's end. But I'm unsure since neither Chrome or Edge (the only other Chromium apps I have available to me) use the same kind of popup for alert or confirm or display this issue.

[codebytere]

Issue opened twice: #41603

[LqdBcnAtWork]

Issue opened twice: #41603

No idea how that happened. Didn't notice until I had left work as well.