Your model is right and the conclusion you're drawing from it is slightly off. A pivot to hold isn't supposed to produce a good rental. It's supposed to produce a survivable one, meaning it stops the bleed on a hard money loan while you wait out a soft resale window. Break-even at 2,900 rent against 2,800 of debt, taxes and insurance is not an investment, and it's also not a forced sale into a bad market. Those are different outcomes.
Where the advice does break is that you've measured the fallback at your all-in cost. Almost nobody's gut deal cash-flows when you refinance out 100 percent of basis, because you built a 465k house and the rent supports maybe 380k of debt. The operators who genuinely have the option available either leave cash in at the refi, pulling 300k instead of 348k so debt service drops, or they bought at a price where all-in was low enough that a 70 percent refi retired the construction loan. Your 320k all-in against 465k ARV is a good flip spread. It is also a thin hold spread.
So the question at acquisition is how much cash you can afford to strand. If you can leave 50k in the deal, your fallback is a modest hold. If every dollar is committed to the next project, your fallback is a sale, and you should be underwriting a lower exit price rather than a rental.
The risk in your numbers I'd push on: you have 22k of carry on a 140 day model and 155k of rehab, which is a 14 percent carry-to-rehab ratio that assumes draws land on schedule. If your lender's construction loan matures at 12 months and you're at month 10 waiting on a certificate of occupancy, extension fees and rate step-ups can hit before any pivot is even available. Get the maturity, extension cost and refinance seasoning requirement in writing from both the bridge lender and the takeout lender before you close, because a seasoning rule that requires six months of ownership post-completion can strand you between the two loans.
