Welcome to the forum CarcinogenX and thanks for getting in with the questions up front.
Now is a great time to explain my previous comments from a programmers' point of view in something like laymans terms...I hope I do it justice.
Firstly, each project contains a weights table that evens out the credit distribution across CPU platforms.
Whilst a project is going through development those weights may be altered to help balance out points allocation.
Some CPUs finish certain types of functions faster than others and that can directly affect the amount of credit that gets claimed for a work unit.
When a project's core client is modified (the bit of code that does the real work - not the BOINC client itself) a chunk of code that gets used often may get changed from being Integer to Float dependent or SSE to SSE2 optimized (for example) and that may change the CPU platform that best suits it.
In Alpha and Beta projects, core clients may get recompiled often so it is hard for the project-lead to keep the weights table updated (more so in the case of an Alpha project).
During this development stage it is the focus of the project team to get the project core-client to perform it's task as well as possible. Later on in Beta they are able to focus more on the BOINC platform balance that will keep most members happy and coming back for more work to do to further the cause.
As the project becomes more stable and the client updates become less frequent, the weights table may be more accurately assessed and updated.
Many of us have (or have had) fairly "interesting" investments in hardware that we have dedicated to our Distributed Computing commitments. Some more than others. Hence it is to be expected that those with the greatest investment will be drawn towards the projects that are past Beta or are at least fairly well advanced in their Beta stages so that the amount of human time and computer time is used to the greatest advantage and an absolute minimum of effort is lost. (Note that - Efficiency does not always equal effectiveness)
Committing to Alpha projects does pose risks to the aim of "recorded" contribution as from time to time mistakes happen and points can be cleared or simply "lost" due to process / hardware / software configuration failure. It happens. Some of us have lost 50K, 100K or more credits in the quest to help test and develop these projects, but doing so knowingly of those risks. Like I said - it happens.
So, when choosing projects, if you have expectations of stability, you should always do a bit of reading first before selecting and jumping onto a project. That way you are less likely to be disappointed.
For me, I like DC'ing on the edge and contributing in the early stages of a project. That's just the way I am.
Well, I guess this is mostly opinion rather than information, but I hope it helps you to achieve your expectations in the DC-quest.
Once again, welcome to the AMD Users forum.
We look forward to getting to know you.