They made tiers, made it simple and easy, and promoted it.
Before the fund really was pushed... they were getting about 5,000 USD per month to develop it... Now it is sitting at 215,000 USD per month.
More money = More developers. More Developers = Better product.
Yes it didnt happen overnight that increase, but it was slow and steady.
>More money = More developers.
There was actually another "kickstart" before that kickstart.
Blender's open-source timeline has a very unusual history in that it was a commercial product funded by €4.5 million in VC capital. Those original investors lost their money by selling in a "down round" to a 2nd set of investors. Those later investors also then lost their money by selling back the source code for a discount of 100k EUR to today's non-profit Blender organization.
https://docs.blender.org/api/htmlI/x115.html
One of the reasons (but not the only reason) that other examples of open-source projects like ... Gimp lagging Photoshop, or FreeCAD not being as polished as Fusion360/SolidWorks ... is those tools never had millions in investor money paying the salaries of 50 developers to kickstart them. E.g. FreeCAD has a non-profit fund but it doesn't attract the same mindshare as Blender did in 2002: https://www.google.com/search?q=freecad+non-profit+fund
Just because Blender started a fund doesn't mean any open-source project can also just "start a development fund" and attract the same level of donations. Blender has some extra history and circumstances in the timeline of "cause & effect" that a random open-source project can't easily replicate.
Blender circa ~2002 had a level of mindshare + evangelism + momentum that most open-source projects do not have. Those ingredients have to be there first to help attract donations to the fund.