This depends on the ionization efficiency and how much of the particular ions are generated. If you are looking at the total ion chromatogram, you can use this as an approximation. 2:1 would be close (and that does not answer the question of how close is close) but you must be collecing across enough of the full mass range to make this assumption.
I have often had a gas chromatogram with some water in it - and I ususally do not collect m/z 18 - but do notice the baseline drop because of ion supression - and the rest of what I am interested in comes off the column. If I had been collecting m/z 18, there whould have been a huge peak, saturating the detector. But I was not collecting the major fragment from water - so I would come up with: no water. Moral to the story - the ions you do not count are not included in the area you use - and if you do not count the most abundant fragmet, you have lost a good part of the signal you intended to use.
Assuming that you are collecting over a wide enough mass range that the total area chrmatogram accounts well for all ions created, you still have a problem. Some compunds ionize more readily than others. It is much easier to force and electron out of a pi cloud than out of a sigma bond. So, if I have equimolar quantities of hexane and benzene - and I get them to the GC column with out discrimination in the inlet (another problem you have to address with quantitation) I know to expect to find the area for the benzene peak to be about 10% larger than the ethane peak.
And, I'll toss one other issue into the mix - Some use the total ion chromatogram as roughtly equivalent to an FID trace. The FID trace is essentially a carbon atom counter - so response is roughly proportial to mass of material present. (This assumption about the FID is a useful generalization - carbon disulfide has no response in the FID, making it a nice solvent for FID work - so beware of the generalization. And the assumption of equivalence of FID and TIC is another useful generlization.)
Bottom line: If you have a 2:1 ratio in peak areas - get standards and build a curve. If you can not get standards full range TIC traces will approximate mass ratios - but you have to be careful with that approximation. The best answer to give is "The two copunds are detectable." If you have to give numbers because you are comparing two samples, the safest to report is that the ratio of compunds A and B is n times larger in sample 2.
