I've always been dubious of all of the reports that say that 5-MeO-MiPT requires double the oral dosage when smoked. I had just kind of suspected that it was poor vaporization technique of fumarate salts that were responsible, so now I'm even more curious.
That seems like the best explanation w.r.t. the difference in potency.
I guess we'd need to know what percent 5-MeO-MiPT goes to 5-HO-MiPT on the first-pass (which could wildly vary, as Kaleida said), but either way the 5-HO-MiPT is going to have poorer absorption than 5-MeO-MiPT and you would need an uptick in efficacy to offset this.
As you can see on pubchem, bufotenine has a XLogP3 of 1.2, which I've been led to believe will contribute to low blood-brain barrier penetrability (which it is definitely known to have regardless). You can compare to the value for serotonin, which also doesn't pass through the blood-brain barrier well, presumably even worse than bufotenine, which is 0.2. By contrast, that of DMT is 2.5, that of psilocin is 2.1, and that of LSD is 3.0. In my observations, extending the bulk of the synthetic tryptamine tail also seems to push the XLogP3 in the same direction; for instance, that of DiPT is 4.1, and that of 4-HO-DiPT is 3.6. Believe it or not, pubchem also has a page characterizing 5-HO-DiPT, and its XLogP3 is a respectable 2.8.
It's worth pointing out that logP applies to unionized molecules by definition, and for molecules which are ionized at physiological pH (e.g. amines in general), logD is a much better proxy for absorption as it incorporates this ionization information (0 < logD < 3 being the criterion for passive brain uptake).
In addition, logP measures the distribution between water and octanol, so it will overestimate the bioavailability of the 4/5-hydroxy-tryptamines relative to the conjugate methoxy (due to H-bonding between the hydroxyl of the tryptamine and that of the octanol). And since logD is derived from logP, it will also overestimate in this regard.
On ChemSpider the predicted logD's of 5-MeO-DiPT and 5-HO-DiPT are 0.84 and 0.46 respectively, so I imagine a similar difference between that of 5-MeO/HO-MiPT (and slightly lower absolute values, due to the shorter N-alkyl chain). For 5-MeO-DiPT this gives ΔlogD=0.38 relative to 5-HO-DiPT, although the bioavailability difference should be slightly greater than would be expected from this alone (due to overestimated bioavailability of the hydroxyl-sub).
I guess this does neglect the possibility of active transport or inhibition of efflux transport, both of which would increase brain uptake. It was recently shown that psilocin is not a substrate for OCT's, so active transport seems unlikely. Interestingly, ketamine is a substrate for OCT's (which are also expressed in nasal epithelium) and imo this explains it's higher nasal bioavailability compared to rectal, which otherwise seems quite paradoxical if you only look at its pKa and pH of the respective tissues.
Even 5% of the 5-HO can color the way the 5-MeO trip would come off if both add up to the same level of activity.
The problem is that the functional assays I've seen don't suggest such a dramatic departure in activity of the 5-HO relative to the 5-MeO. For example, Blough et al. found calcium release EC50 of 3.49nM and 3.87nM for bufotenine and 5-MeO-DMT, respectively. Perhaps if the functional selectivity of the 5-HO differs from the 5-MeO you wouldn't need much to color the trip, but unfortunately the above study didn't measure arrestin recruitment for bufotenine so it's difficult to speculate in that regard.
It would really help towards resolving this question if we had some higher-level efficacy measures (i.e. layer V electrophysiology) of these compounds, but I'm not sure even bufotenine and 5-MeO-DMT have been studied in that regard.
That's not consistent with my experience at all; all of the synthetic 5-methoxytryptamines I've smoked start kicking in almost immediately similarly to smoking base tryptamines. Their character does change over time though, at first feeling rushy and flooring again similarly to smoking something like DMT, and also sometimes having some of their strongest visuals with eyes closed during this period too, but after around half an hour I do find them to become more allowing and similar to something like LSD.
The qualitative difference with different ROAs is super fascinating, although I think a lot of it can be explained w/ kinetics. Even if two different ROAs reach the same peak brain concentration, the one which reaches it faster will outrun desensitization/downregulation to some extent and this would lead to stronger peak activation of intracellular pathways. In cases like this where the pathways affect gene expression, these kinetic differences could then crystallize into structural/functional differences which would then feedback on the subjective nature of the trip.
It seems that if the 5-HO was that much more potent the experience with smoking it would slightly intensify as it progressed (as more and more of the 5-MeO was converted to the 5-HO), but I guess that may be difficult for you to compare given the changing qualitative nature of the trip. It also seems like the duration of smoking it would be slightly lengthened as well (assuming the metabolism is rate-limited by liver enzymes), which I guess could help explain the allegedly reduced potency of smoked vs oral (again assuming the 5-HO is more potent).