You have a good point. Household or other-purpose (such as automotive) chemicals can wind up interfering with chemical syntheses or workups in ways that aren’t expected: it just isn’t worth the trouble saved, because when trouble is caused it can be a long time before the problem is pinpointed as being that product.
For example, I saved myself time once (I thought) by using supermarket sodium bicarbonate instead of reagent grade. Only after more time than I’d saved had passed with things going wrong did I discover that Arm and Hammer sodium bicarbonate is very, very impure and was the sole cause of my problems.
I do agree your method has merit. You got rid of far more than 75% of the estradiol benzoate. E2 is an extremely potent substance.
If you had 25% of what was in the original product you’d be a Hooters girl by now 
A good thing would be to take some – even a very small amount – of product produced by using minimal methanol and then see what happens when dissolved into straight vegetable oil at 50 mg/mL, which is the solubility limit of TP in the oil.
If not everything dissolves, then in fact adding the methanol step didn’t reduce the proportion of EB as much as occurs with simply dissolving in vegetable oil, but makes the filtration of the oil vastly easier and therefore is highly worthwhile.
Or if everything dissolves, then most likely (though by chance solubility ratios could be exactly equal but that is relatively unlikely) the methanol step reduced the proportion of EB. And filtration came out better.
So either way, your method is an improvement.
There is nothing wasteful about using “too little” methanol because it is not as if the part retained by the filter must be thrown away.
If after complete drying the remaining weight shows that more grams of material remain than can possibly be accounted for by EB, then it is known that a second pass with more methanol can get some planned portion of the remainder, and from the first pass it will be possible to estimate that amount fairly closely.
I would avoid using added heat at any step of this operation because:
- It gives inconsistent and irreproducible results.
- It is totally unnecessary
- While in your case I don’t have a concern that you will do it in a dangerous way, in any large set of people, if they use heat with methanol some of them will have a disaster.
Anyway adding heat is totally unnecessary and can yield variable results. The only ways to be consistent are to either use room temperature; or reflux (operate at boiling temperature, which is inappropriate here); or use a thermostatted system with stirring. There is no reason to bother with the latter for this particular workup.
If being short on patience and wanting to use heat for that reason to enable faster dissolution, the thing to do to get reproducible results is to allow to cool to room temperature and allow sufficient time for excess amount (that beyond the solubility limit at room temperature) to crystallize back out.
That should also be done if using a sonicator to speed dissolution.