Key findings of my taste selection confronted with my growth/plant health assessment of individual plants, as done at harvest, with indicators gathered in a spreadsheet including pictures, keynotes (using Kobotoolbox on a tablet) :
Taste selection was undertaken on an original 500kg pile of moschatas, stored on shelves in an open barn bay, in really suboptimal storage conditions, slightly above freezing but still exposed to atmospheric humidity. Selection done at 3 months post harvest.
First step was to discard every fruits having the single light storage problem, which would have made the fruit collapse around January/February. So that’s a selection consistant with a long-term storage and use objective. So 35-40% of the fruits were gone by then.
Then taste selection was conducted in two steps:
- First : raw step, using a spoon, to pre-select the best, looking for sweetness AND aromas (things like a pronounced buttery flavor)
- Second : baking step, in thin layers in a pan (see above post), to confirm that first step, precise it. Helped also to discard about 10% of “false positives” (= falsely “great tasting”, as tasted raw)
The really really really interesting thing was to confront that sensory evaluation at 3 month post harvest of squash stored in suboptimal conditions WITH GROWTH INDICATORS I had gathered :
Here are the key take aways :
- Zero plant with bad to average growth indicators did produce a single remarkably tasting fruit. So from those struggling plants, which accounted for 40 to 50% of the remaining squashes, absolutely none made a remarkable fruit.
- All selected for very good taste had great to exceptional growth indicators, in particular the “over the top” taste selected : all high to super high yielding, all having long vines, most showing exceptional rooting and rerooting capacities, most still growing at harvest, some even flowering again! Other way to say it : none of those fruits were picked up on collapsing plants at harvest.
Meaning to me and for practical purposes : growth indicators, plant health indicators, can be consistantly used as the minimum threshold for selection not only on “adaptation” but ALSO on great taste, which is a fantastic take away! 
Those - or at least in my place and in moschatas - seem to be formidable proxies :
Consequentially :
-
At harvest, that means that I could focus on storing properly the fruits from the 10-25% most exceptional plants, and so sell, give, or eat the others, without the burden of looking after them, storing them in a proper manner. All my taste selection would be focused on those great individuals, so dimishing a lot the taste selection work.
-
More than ever I’ll maintain my early vigor selection, but as much as possible late in season, so with bigger spacings : to see more accurately plant development before culling the slow growing, those less “adapted”. I would love to go as far as 45 days post sowing to do the culling, and maybe in 2 culling steps.
I’ll just add up to that that “high yield” apart hasn’t been a great proxy for great taste, and nor for long storage. That is consistant with what Brent Loy already outlined in his “Grower’s Paradox” document : breeding for yield only is kind of easy… but it doesn’t tell you anything about…quality! As in squash, long storage necessitates high levels of starch at harvest, so to say superior photosynthetic and metabolism capacities, as starch is a complex molecule, much much more complex than the simple sugars. which still can make a decent squash at harvest but if not balanced with high starch level, will make fruits collapse early… And as some say "winter has been bread out of winter squash”, we can guess how…
Also, and as outlined by the Kempf-White-Mella trio, plants with superior health, superior soil-plant interactions (like in my squash case : vine long length, rerooting capacities, leathery texture of leaves, intense greens, continuous growth super late in season, and new flowers, and giant tendrils…) are more inclined to create aromatic compounds, a.k.a. secondary metabolites. This also has been relatively consistant with my taste selection.
That’s why I say that plant health is a really good minimum threshold for selection. And actually it’s what farmers, humans have been doing forever… 
To be continued…