@gbathree – I would love to explore this! 
My impression is that the materials for everything you described (except the soil moisture sensor part) – could probably sourced for < $50 currently – certainly if one dug deep into the supply chain / did some research on Ebay / etc; and absolutely if one arranged relationships with manufacturers / factories. (But honestly I think you know the hardware supply chain world better than I!)
The reason I except the soil moisture sensor is that my impression is that while it’s easy enough to make a cheap capacitive sensor that reacts to immediate, severe changes in moisture (think: dry potted soil suddenly getting watered), these ‘cheap’ soil moisture sensors (typically capacitive) don’t give very meaningful measurements over the long term when used to assess soil moisture in the field, because these sensors tend to be so heavily impacted by non-moisture parameters – highly localized variations in soil type / composition, in particular.
I believe (would like to verify!) that there is a lot prior art trying to calibrate out / account for this variation, with the idea that many, distributed, inexpensive soil moisture sensors, in combination with some e.g. a few more sophisticated sensors, and some modeling, might still yield meaningful results; I’ve heard one soil physicist who had reviewed such efforts characterizing it as a ‘fool’s errand’.
To make things worse: validating the sensor’s performance is difficult – typically, to compare apples to apples, one needs to meticulously remove soil samples around the sensor, weight them, bake them, weight them again, and replace the relevant soil type back around the sensor. This means we end up with a lot of Ag startups selling the cheap soil moisture sensors, and consumers thinking that they’re getting something meaningful (“the reading spiked when I poured water on it!”) without anyone in the loop doing the due diligence to validate the sensor tech.
I do want to maintain an open mind about the idea, of course; and perhaps there are use-cases (yours?) where cheap soil moisture sensors would provide meaningful insights (what information do you hope to glean from them?) …
This all makes me think that it’d be nice to organize an effort to gather up relevant papers / reviews out there on the topic of ‘using cheap soil moisture sensors for X’, and see what the current consensus is. The result of such of GOAT snapshot review of the soil mositure sensor literature would be really useful to publish for the wider community to see; either because there are useful ways of deploying cheap soil moisture sensors (and we should jam on making some open hardware versions of them!) – or there aren’t (and people shouldn’t waste their time / money on them!)
Cheerio!
DB
