i spent coins early on whenever i had the chance just to support the ecosystem and support vendors who accepted btc. at todays prices i dont even want to think about the value of the coins i more or less blew just to support the ecosystem, but i dont regret it a bit. after all its we who are proving the value of btc. we need to prove it "just works." to me that btc was well spent.
Regarding your substantive point, cited above, I have pretty much maintained a system in which I replace any bitcoins that I spend, and even though I don't go out of my way to spend bitcoins, if I find an opportunity to spend them that does not seem to be too much of a burden, then will soon thereafter replace any spent bitcoins. I was a lot more adamant and nervous to replace them right away in the early 2014 to the early-to-mid 2016 time period when I was engaging in a lot more BTC accumulation (largely establishing my BTC postion - at least up to late 2014 and a kind of maintenance of BTC holdings thereafter).
These days a remain a lot less nervous about immediately replacing any BTC that I might spend, but I still attempt to reasonably plug the dollar amounts of any spent coins into my authorized buy back amounts. So for example, let's say that I go out and I see a 5% discount on a $1,000 product, if I buy with bitcoins (only way to get the discount is with bitcoins). I would be like "wow", that seems like a decent enough incentive for me to spend some bitcoins. Therefore, I spend $1,000 in bitcoins to buy the product, with a $6,300 exchange rate. (actually, 5% means that the seller is giving me the equivalent of $315 extra for my BTC ($6,300 x .05), which is $6,615.
In the 2014-2016 period, I would nearly immediately lock in the 5% profits by buying back BTC immediately around $6,300. These days, I might let the profits ride a little bit. Therefore with the $1,000, for example, I might set buy-back orders at $100 increments, or some other reasonable ladder down the chain.. So maybe I conclude that $6700 is about as low as is reasonable in our current BTC trading range so I would set about 6 buy orders of a bit larger than $1,000 ($1,000/6 = $167)... maybe around $200 for each buy back order at $6,200, $6,100, $6,000, $5,900, $5,800 and $5,700) .. so if all of the orders fill, I end up buying back around $1,200 worth of BTC for the $1,000 that I spent. One other tip is that I would not set my buy back orders exactly at the round numbers of the $100 increments, but instead set the orders a bit above the $100 increments in order to increase the likelihood that they will fill and not reverse just prior to filling at the round numbers where everyone (including BIG whales) tend to quickly set their attempts at BTC price manipulation orders.
i had been a btc miner (around 2011-2013 or so when gpus could do it), not a buyer, of btc, although now i occasionally also buy some. so it was more or less like continuous dollar cost averaging buying. so the coins i mined back in the day (which is the bulk of my current stack) i never replenished as i spent them back then. i mined new ones anyway so i was spending profit, and just wanted to test and use the new ecosystem.
replenishing coins sold with new ones is the best way for most people looking to build up their stash while taking advantage of btcs strengths for purchases. your strategy is a good one. currently i mine alts with gpus and exchange them for btc so i still replenish btc although its not much.