Ugh, I think the new UI is terrible. Too much white space, buttons in random places, no natural flow to this tool.
A big part of my job is user experience. The new UI is significant
superior to the old UI in at least one of the specific ways you've mentioned. The old UI was far too crowded. Most human beings need the whitespace to help them better process what they're seeing. Also, other than putting the data in (see below), there is a pretty decent natural flow. The flow seems to be optimized to support review of the analysis, not data entry. My own experience is that trying to achieve two different consumption objectives with the same UI is bad UX. My own software does the same thing - it was just far too expensive to consider the correct answer: Creating two independent UIs, one for data entry and one for review, analysis and such.
There are some unfortunate button placement choices, but I'm not sure yet whether they're actually misplaced or if instead I'm allowing my familiarity with the old tool to color my perception. More specifically, the Accounts & Income Sources button is very poorly placed. Personally, I think there should be a tab prior to "Savings Rate" that lists "Accounts & Income Sources", with a button on that tab to add/remove/change the data. I also think that the Hypothetical Detailed Income/Assets drop-down and the Underperforming Market/Average Market drop-down should be alongside the gear and show tabular data hotspots in the chart, the gear hotspot really should be a "Today's Dollars/Future Dollars" drop-down to match the drop-downs I'd put aside it. I also think the tool, itself, is buried too far deeply into the Fidelity website structure. Other than that, I believe Fidelity made substantially correct UX choices.
Now: What does it mean if you don't agree with my criticisms of the UX? Even though I'm trained in this stuff, and do that work on a daily basis, I'm not willing to say "I'm right and you're wrong." Most likely, it means that we're both wrong and Fidelity got it right. A lot of this work involves applying some basic theories (as I alluded to above) but then watching a lot of real people of different sorts using the system, and learning from things as subtle as what order their eyes move about the page (much less what buttons they try in what order...)
Would have been much better with some sort of side menu so you can easily move around to change parameters.
That's an idea. The old RIP didn't do a good job in that regard either.