It's important to first think of the active system at play and its influences on people, and second, try to improve it without disconnecting the model from reality.
This is a hard problem, and is one of the reasons leaders and managers need systems thinking skills above all.
[Citation needed]
Also, I find the version with paperwork better, because that creates a documentation trail. If someone is routinely favoring certain people or groups, documentation makes that easy to expose. As opposed to the "word of mouth" system where it's likely impossible.
Ability to understand your specific system and guide it is the far more important factor.
You ask for a citation for an opinion (see: "preferable") then proceed to share your own semi-claim without a citation.
Though probably not intended, saying "Citation Needed" can be off-putting and wastes time in discussions where if you spent a few extra seconds, you could ask a more thoughtful question. For example, your abridged quote has numerous elements and the target of your critique is unclear.
Lighten up!
> Both systems have their downsides. The simpler system generally has fewer...
I merely quoted the remainder to complete the sentence.
It appears simpler at first glance, because the complexity is deliberately hidden. The oft-quoted example concerns “structureless” organizations:
http://www.jofreeman.com/joreen/tyranny.htm
This is not universal, but something to guard against: That the “simpler” system is not, in fact, a structureless environment where manipulative individuals will create an invisible complex system.
That sounds very much like "hiring people based on having a reasonable conversation about their experience.” It is a perfectly sane and reasonable system, provided everyone involved is acting in good faith, has a great deal of ability to evaluate people based on their actual contribution to the success of the company, and nobody has any overt or internalized biases.
Without taking anything away from your company, I’m sure you will appreciate that in many companies, this business of “assigning you a bonus based on conversations with the project manager(s)” gets twisted into a horrible political game where people build fiefdoms based on tit-for-tat agreements, and where all sorts of biases come into play, e.g. the men are all “take-charge go-getters,” while the women are all “confrontational and bossy.”
I’m saying nothing about performance evaluations being better, just that results like you have experienced with your company are remarkably difficult to reproduce and sustain.
My bet is that if conversation works in your company, performance evaluations would also work. Maybe with more annoying paperwork involved, but it would still work.