There are several reasons that can result in inconsistent results for "halflife."
The terminal half-life and the elimination half-life are two separate things. One measures how long it takes to achieve 50% elimination after peak, the other measures how long to achieve 50% after the final "straight line" slope is achieved (the 4th/5th halflife typically).
These two are drastically different, and the terminal can be more influenced by the type of metabolism pharmacokinetics (single compartment vs 2-compartment 2CPK or multicompartmemt model).
The 3D shape and mass of a molecule also influences its metabolism and excretion, so test phenylpropionate and nandrolone phenylpropionate don't actually have the exact same half-lives based on solely their esters. Nandrolone has I believe a smaller molar mass than testosterone, so it would in theory clear faster, but when attached to phenylpropionate ester chain NPP Is reportedly slower than TPP. Another example of test enanthate and tren enanthate. I think there are studies showing TE with an elimination half-life as fast as 4 days, and TrE something like 7 days.
And completely separate from the chemistry of molecules, every human has different biochemistry that'll influence how fast/slow we metabolize drugs. Enzymes, vascularization of administered tissue, metabolic health, genetics, etc. So two guys that are both 30 years old 180cm 100kg 10% body fat and identical twins can metabolize 200mg Test Enanthate differently depending on lifestyle, diet, sleep, etc. Create a study with hundreds or thousands of people, the results will be all over the place.
Then the studies become the issue. The testing methodology, equipment used, other parameters that can influence testing results, etc. Just the other day I saw a study about halotestin having a half-life of 8 hours, but most studies say it's 2 hours. The methodology for determining the half-life in the 8 hour study didn't use just blood levels of halo, but also factored in how much metabolites were excreted in urine and the density/rate of excretion. Which again can be heavily influenced by that person's liver/kidney health, hydration status, stimulant use, general electrolyte balance, diuresis tendencies, bladder capacity, stress, etc.
I do not think it's possible to say "this molecule's half life is x hours," because it can vary so much. But I do think it's safe to say averages [within standardized testing parameters] can exist. Like we can all agree test prop peaks and clears much faster than test undecanoate.