Promethease does rank snps based on research and gives links, plus it is searchable for any gene you may be interested in and is very comprehensive. Some people do find it difficult to follow unfortunately.
As
@Oci notes, MTHFRsupport is fairly comprehensive and has good diagrams showing where the various genes fit into the relevant metabolic pathways.
You have to realise however that just because a snp is flagged doesn't mean it has much consequence. On the MTHFRsupport website they do provide some links to SNPedia which might help in evaluating what the snp means (sometimes though SNPedia is in error!), but for many snps there is no research at all, or maybe just a single dubious association study.
You really have to do some work of your own in deciding whether or not the flagged snps might have consequences. Some snps are discussed on PR, otherwise it's just googling and trying to find the research behind the claims.
Once you understand a bit about the terminology, common sense is also a help. A single isolated +/- snp is highly unlikely to mean anything at all.
A lot of +/- on a particular gene might be more significant, particularly if they were all on the same DNA strand (so-called compound heterozygous), but there is no way of knowing if this is the case since 23andme doesn't record which strand is associated with which snp. Still several +/- would be worth following up.
A single +/+ on a gene might be more significant than +/- but it entirely depends on the snp. You would need to check it out. Some changes have no effect. Several +/+ should make you thoughtful and would certainly be worth following up, but again, it depends on what the changes are.
As a general rule, a change in the part of the gene coding for protein is more likely to have an effect than changes in other parts of the gene, but not always. An altered protein with altered function (either slower or faster than normal, or perhaps more unstable than normal) could be made and this could have consequences. You would need to look at the research to know how much of an effect this could be.
Some changes in coding region are silent - ie the DNA is changed but no change in protein results. This is the case with the MAOA R297R snp which I discussed above. These types of snps mean nothing in isolation. Possibly in combination with a number of other snps there could be an effect but you would need to read the research.
Sometimes a snp upstream of the coding region in the promoter region (which switches on transcription of the gene) can affect the rate at which the protein is made - ie there could be more or less of it.
Snps in other non-coding regions are highly unlikely to have any effect in isolation, but sometimes a lot of them can make mRNA unstable and so affect the amount of protein made.
You can find out where snps are located in a gene by searching on
dsSNPs.
Once you've done these sorts of assessments, then you have to make up your own mind about what to do. As an example, I have several +/+ snps on the gene for the enzyme that converts beta carotene to vitamin A. Some of these are known to adversely affect the function of the enzyme - not dramatically, but an effect. I therefore think it is wise for me to supplement vit A.
Similarly I have a number of +/+ snps on the fatty acid desaturase enzyme which makes longer chain unsaturated fatty acids from linolenic and linoleic acid precursors. Again at least some of them are known to adversely affect enzyme function, so I supplement borage and fish oils.
On the other hand, I have quite a few +/- MAOA snps which I have concluded mean very little and quite a few +/+ COMT snps which I have concluded probably have a minor effect. I don't take any particular action for either of these (though I do supplement their cofactors for other more general reasons), I certainly don't give any credence to the myth that COMT+/+ can't tolerate methyl groups - I have no trouble with them.
Sorry there's no easy answer, you have to work at it.