Though one oddity, in some paperwork I've seen they're planning on 2 M4-TL-US inverters and 1 M8-TL-US inverter. The M8 I would figure for the 6.7 and .9 KW section (since they're facing the same direction), but the M4 seems undersized for at least the 5.6KW south-southeast section there. I've asked the project advisor if that's actually correct.
DC/AC ratio is a hot topic with lots of data sources and theories, but I think given the high operating temps of the solar roof it makes sense for you to just keep the 4kW Delta inverter for that group of tiles
Here's an article from Solar Power World that says for string inverters, the target DC AC ratio is 1.3 to 1.6. This takes into account the financial implications offset by clipping. So I think what Tesla quoted you for that M4 (5.6 / 4) = 1.4 should be ok.
https://new.abb.com/docs/librariesp...ded that an inverter,fixed cost of the system.
ABB found that the optimal DC-to-AC ratio in these areas depended on the design goal. For example, regardless of site conditions, sizing a system to maximize specific yield allows for an ideal DC-to-AC ratio at or slightly below 1.2. However, sizing a system to target the best financial output could lead to higher DC-to-AC ratios, between 1.3 and 1.6
The article cites research by ABB (who subsequently bailed on the inverter business so maybe they were trying to sabotage? haha)
One thing I'm curious about with the Tesla solar roofs is why they can't add an extra row of tiles near the ridge of a roofline. Its there some huge setback requirement that they need to comply with? Every time I see these Tesla solar roof layouts I'm always thinking "why don't they fit some more tiles??"