OK, some comments:
The reduction of azobenzenes shouldn´t be a problem, i have some procedures around, using SnCl2 and sodium bissulfite, i think. If you want me to post them just ask.
As to the rest of your idea I think that the aniline group would react with DXM in the methylenation reaction so maybe the way to go is the sandmeyer reaction.
All this sounds like it´s too much work just to get to bromobenzodioxole, wich would require a grignard reaction with allyl halide to go somewhere interesting.
And the biggest problem with the route from vanillin or eugenol wich, for the ones without access to DMSO,DMF or PTC´s, is the methylenation is still required in the route you propose.
Personally I think this might be a better OTC way to benzoquinone via aniline or other subtituted anilines and KCr2O7 than to safrole. But If you have all the necessary reagents for the methylenation then nevermind my ramblings and just go on with it.
;D
EDIT: Maybe you can forget the sandmeyer and use this procedure http://www.orgsyn.org/orgsyn/prep.asp?prep=cv5p0139 (http://www.orgsyn.org/orgsyn/prep.asp?prep=cv5p0139)
(referenced from rhodium´s page) to go straight from 3,4-dihidroxyaniline to protocatechuic aldehyde.
you are suggesting to transformate amino group to aldehide via diazonium ione , and after that methylenation of product to get piperonal
Not necessarily. I think you can use the sandmeyer reaction, as you said in your first post, to get bromobenzodioxole. But, without an allyl halide you can´t do anything with it(unless I am forgetting something).
From piperonal there´s also the possibility of using the darzen´s condensation(or whatever it´s called) posted by Barium.
For OTC preparation of allyl iodide you can try this: Post 357095 (https://www.thevespiary.org/talk/index.php?topic=6766.msg35709500#msg35709500)
(Captain_Mission: "somewhat off-topic...", Chemistry Discourse)
I´ll be back in a few days with the reduction procedures.