That's really a sticky wicket.  You're looking at low permeability drugs (either class 3 or 4).  for oral route, that means you're likely pinning your hopes on small intestines in a fed state. I believe there are emulsions and amphoteric molecules that can improve availability.  That said, all of the approximations made by simulation or by experiment usually don't yield satisfactory answers.  For anything better than a magnitude of order (at best) you're probably looking at animal studies.

Comparing different routes based on different experiments sounds attractive, but then you observe that a specific test method (dissolution) by itself is fickle and inconsistent.  Pushing out into places where different routes are explored via in vitro or in silico is a strange forest where little is known.

That said, there are software packages that simulate these interactions.  There is possibly one that is free for academic use.  I'd also recommend reading everything you can on IVIC.  

Good luck.

 If first pass metabolism affect oral administration then try parental route for adminstration. 

Helena, the only way I have found to accomplish what you desire is by pharmacokinetic evaluation using complete mass balances for the parent and major metabolites after a subject is at steady state.

HPLC is the best methodology as it deals with parent drug, phase I metabolites, and phase II conjugates.

Normally, only plasma and urine is needed.  Plasma sampling with time is rather straightforward, urine sampling is more complicated as it: [1] requires collection of all of the urine over the test period and [2] adjustments for the pH of each urine sample based on at least the H-H Equation (which is only an approximation).

Generally, I find it best to normalize the urine samples for dilution and pH--creatinine indexing is very inaccurate compared to refractometry.

I use to do the above with opioids in my laboratory many years back.

Email me if you need more guidance.

This is a  general question and as mentioned before there is no easy answer. The answer would depend on several aspects. For example is this an academic exercise or a drug discovery project? Second how many compounds are we talking about? Third what are you going to use the data for? Also when you say lypophilic compounds, there is a wide range of lipophilicity. As a general rule, again talking about oral absorption to narrow the discussion down, two factor, in general, that influence oral absorption. First is permeability and second is solubility. Usually compounds with log P of 1-3 has better permeabilty although there are some exceptions. There are software that can usually predict these two parameters fairly well. But you have to do some studies on a few compounds to confirm the prediction. Now bioavailabilty is a composite of oral absorption and first pass effect. The compound could be well absorbed but shows poor oral bioavailabily. You can get an idea about first pass by determining the metabolic stability using in vitro system such as liver microsomes or hepatocytes, although IV clearance in animal is the best estimate of first pass effect. I would suggest that the compounds that shows the best permeability and solubility and low metabolic stability can be tested in a PO/IV study in animals (usually rats or mice). When you do the PO study it will be best to dose it as a solution to avoid compromising bioavailability because of low solubility. There are many relatively safe vehicles you can use to solubilize  hydrophobic compounds (Gad et al., International Journal of Toxicology 2016). One of them I used a lot is 20% aqueous hydroxypropyl beta cyclodextrin for both IV and oral administration..Hopefully this will help