There are many people in the community divided between "Dividing rewards equally among projects it's a good thing as less exciting (but not necessarily less useful) projects can get more computational power" (as i believe), or that more popular projects should have bigger, proportional rewards. The solution i bring here should appease everyone (or satisfy no one).
Lets take a look. At the moment the system works like this : 
Every project gets their piece of cake, and its nice, but it also means some frustration due to the lack of profitability of a lot of the most interesting projects, like WCG or SETI@home. Plus more pressure to whitelist projects that may not pass the cut, like the whole MooWrapper issue. I also have some problems with Amicable, but the dataset may have future applications, so at least qualifies at science.
I propose two alternative similar solutions :
My favourite one is this one : 
How does it work?
We have 3 basic categories, with a few example projects
- Gridcoin Maths
- Yafu
- SrBase
- PrimeGrid
- Gridcoin Physics
- Universe@home
- Einstein@home
- LHC@home
- Gridcoin Chemistry & Biology
- Tn-Grid
- Rosetta@home
- DrugDiscovery@home
And the most interesting one, which can be the more popular project, or the most popular of each category, i would go with the former. (It also needs a better name)
- Gridcoin Top
- Seti@home
- WCG
- Lhc@home
Lets do some math.
Under the current model every project gets a 3.84% of the GRC, (100/26) , with the current system, a Gridcoin Maths project would receive a 2.27% (25/11 [counting enigma as a math project]), they take a hit, but they are the less popular projects, and also the ones who run most nicely in the most powerful cpus and gpus.
A Gridcoin Phyisics would receive a 3.15% (25/8) . A small loss, but it also includes much less popular projects like sourcefinder or Milkyway@Home . Profits on projects such as Einstein@Home or Universe@Home would probably stay the same.
A Gridcoin C&B gets a substantial boost on profitability, well deserved as they tend to be the most popular ones, and most likely useful in the short term, with a 5% of profitability (25/5, counting CSG as one, even if it has a math subproject)
As for the Gridcoin top, my personal judgment is that there are currently only 2 projects worthy of being included, WCG and SETI@home, others may not agree, but i'm just giving ideas. They get to multiply their share of the cake for almost 4x, with a 12.5% , but they probably will still not be the most profitable ones to crunch for. Hopefully this could attract major scope projects to come to the boinc network with GRC workforce as an incentive.
An alternative way of implementing the same thing, which i think will be much more complex would be to reserve part of the bounty and distribute it to the most popular project of each category
I also think it would be unfair, Primegrid or Collazt is for sure not as popular as WCG or S@H.
The main reason for which i would like to see this project approved is to attract the huge amount of casual WCG and Seti@home crunchers to our blockchain.