I think the jury on that one is still out until we get the long awaited anti-flicker update.
OP, Xojo:
can you change this subject to something else (something that not let people think Oh ? 2017r3 was released ?"
TIA.
Yes, when I saw this thread title I open Xojo Download page. But, I just closed it, it’s not a big problem.
How about a stable version that gets bug fixes for 1-2 years and a bleeding edge version that has everything?
That way people could stick with something that works for them and importantly receives bug fixes which isn’t the case now whilst starting new projects with the latest and greatest version?
This would be a win win for Xojo as subscribers would be incentivised to stay subscribed as they would be getting bug fixes for the version they are actually using which isnt the case right now. This in turn could lead to a more stable cash flow for Xojo and more developers to fix/add stuff (yay) 
Then scrap the current update method of downloading an entire new version and make it into an differential updater that gets the changed files only. Only have new installs for major editions.
Thank you.
At this point, for some reasons sometimes you HAVE TO update in order to benefit a fix for your project.
But the dilemma is always that with the update you also get stuff you don’t like because it’s will decrease the user experience of your application. At the moment I am using the newest IDE for Windows projects, but mostly use 2016R3 for building, which results in less flicker etc. Of course I want to go for newest techniques but sometimes the implementation needs one or more releases to get equal or even better results than what you already had.
As of now, we don’t even have a chance to get things fixed in a previous version, even though migrating to the newest release is sometimes not possible for an existing project.
I would say, if you archive a project, archive the used xojo version and plugins with it.