I'll give you "better R users." However, I think "better R programmers" is up for debate. I'm not opposed to people programming with tidy evaluation, but it isn't all roses and rainbows. It has a performance cost. How much that performance cost affects your decisions depends on how you envision your work being used. If it gets used once, probably not a big deal. If it gets used in any kind of resampling, it can be a very big deal.
For instance, using the examples here, avoiding quosures altogether results in an execution time of about 200 microseconds. Using the quosures takes about 20,000 microseconds. If I needed to run this in a routine 10,000 times (say for a bootstrap procedure), that translates into 2 seconds with standard evaluation and over 3 minutes with quosures. There's a pretty good thread on the subject here