So... not gonna really dig into any arguments for or against here, just going to put out the how and why.
IN 1.x you were stuck with overlapping windows, this was a terrible experience and meant you only ever had a couple of scripts and objects open at once - max. Any more, and they were utterly lost in a sea of over lapping madness. I constantly found myself having to close everything to find anything. This sucked.
The workspace lets you lay everything out, or at the least, lets you open much more than 1.x could ever manage and still find your way around. Even if you only opened 3 objects (which isn't much by 2.x standards) you were still ahead. I usually have 10 or more objects open and then use the scroll wheel to shift them all up and down. This started out as "zoom" on the scroll wheel, but this was a horrible experience and we swapped it to scrolling. Sine they open up on top of each other, a vertical scroll is very natural, even if you only have 2 objects open, they are laid out clearly and easy to get to.
The code editor can either be window based, or fullscreen based. If you want to maximise your workspace, then putting code into a fullscreen tab and Using F12 (or the icon at the top) to hide the docks really gives you as much space as you need. If you do find you need more, then ripping these workspace tabs out and maximising your monitors lets you spread your work around.
While the resource tree does work as a "go to" resource we have added in shortcut navigation using the cursor keys, bookmarks so you can bookmark anything at any time, and the ALT-T thing to jump directly. All this is WAY more than 1.x ever had, and again after you opened a couple of windows, you were just stuffed.
The IDE is utterly customisation to us, it's written from the ground up and we can change anything in it if we feel the need is there. Currently, I'm not persuaded by the 1.x window on window argument. I used it extensively to build projects and it showed up windows my annoying flaw. Interestingly, windows 10 has also moved into this "workspace" idea (as Mac and Linux have had for years) with it's multiple desktop. They recognise - as we do, that 1,000,000 windows on top of each other is just horrible. Multiple desktops let you put apps that your are working with into logical spaces together. I have my browser, email and the like on one, then VS, source control on another .
You can do just the same with 2.x using workspace tabs, you don't have to pack them all onto one space and scroll around. stick things that youwork side by side with on the same tab an either flick between them or rip them off onto another monitor. I'd also encourage you to try the fullscreen code editors - and the various preferences to adjust your workflow to suit.
Changing workflow is hard, and I'm under no illusions that everyone will come to like it - just like not everyone liked the dated 1.x workflow. And while I will always monitor how users are doing, and if there is anything we can do to smooth things out, I'm pretty resistant to going back to 1.x overlapping window hell. But, find a middle ground, and I'm all ears.
Lastly... the room editor is kind of a half way house right now. Having parts of it in the dock is not the final layout, but we didn't get time to fully complete it. This will be finished over time (as we do everything else), and that will also help maximise space.
Actually...really lastly. We also have a new preference coming, DPI override. This lets you shrink or grow the layout to suit, and while it's been added as an accessibility aid (for those who struggle to see everything clearly), I've actually come to love it when I shrink everything - just a bit. Normal (100%) DPI is 96, and 75% is 72. I now sit mine around 80 and find this is a really nice look. We may well be able to do a skin that helps shrink the chrome and gives this kind of effect - although the DPI override does do this nicely.
So...as I said, just my how and why....