I still want to know what you mean by work (long term) because 2e lasted long term and 3e/4e book a month lasted long term and yet somehow a book every 6 months, which for the record has not lasted long term, is now the new success story.
2E did not work long-term. The first and most crucial requirement for "working" is that the company stays in business. If the company was reasonably healthy at the start of the edition, and had to be rescued from bankruptcy by the end, that edition did not "work." It failed - the kind of failure that leads to sudden respiratory ailments if your boss is a Sith. And people involved have confirmed that the frantic pace of new releases was central to 2E's failure: TSR pumped out far more product than the market could bear.
3E slowed down the publishing schedule dramatically, and was indisputably a success. At the same time, its success depended on much faster edition turnover. 3.5E may not have been a full edition, but it gave Wizards a way to reprint the core books and issue a slew of new splatbooks. If we count 3.5E as half an edition, then the lifespan of an edition putting out 1 book/month is about 5-6 years.
4E had... issues... but I don't think we can blame 4E's woes on its release schedule. In any event, its 6-year lifespan (or 4-year if you count Essentials as another half edition) is consistent with what we saw in 3E.
I don't think the relationship of release schedule to edition lifespan is linear - 5E is not going to last for 20+ years - but there surely
is a relationship. Splatbooks and supplements don't sell nearly as well as the PHB, DMG, and MM. Those three books are D&D's bread and butter, which means that profits start high with the release of a new edition (when everyone rushes out to buy them) and taper off over time (once everyone has them). Putting out a lot of supplements requires a large staff, which means WotC needs that infusion of cash more often.
So, it's not really about what "works," so much as about what kind of an edition cycle you want. If you're fine with a five- to six-year cycle, then the 3E/4E release schedule is great. If you prefer a little more stability, then a slower schedule is called for. 2E demonstrates that there is, in fact, such a thing as releasing too much content for
any edition cycle, but no post-TSR edition has approached that threshold that we know of.