The real problem with trade

I disagree - the problem with trade is simply that is hands down less effective than making your own items. Construction Infrastructure is cheap and relatively quick - trade is expensive and extremely long.

A lot of people have made suggestions trying to solve the problem where trade seems to be dead, and most are blaming lack of sufficiently large trade centers, travel time, inability to determine the value of resource types, etc.

That last one is close, but still not quite it.  It's the value of gold itself.

The basis of currency is that some resource have a stable value against which all other fluctuating values can be compared and traded.  Gold intuitively makes sense but is actually utterly useless as a base of trade.  All the resources are infinitely renewable, but gold is the fastest growing of them.  Gold income increases almost exponentially compared to the other resources as population rises, putting "inflation" through the roof.

In order for trade to have a base currency, that currency needs to be limited to a maximum quantity based on total world population.  It can't be something that as many players as wish it can just endlessly accrue faster than they can throw it away on blow and hookers.  For gold to have value, it must have constraints that conform it to the traditional 0-sum game, where one man's success must be connected to another's failure.  To establish that, spent gold has to go somewhere, and that somewhere has to be able to run dry when its supposed to come back.

Such a system would not be computationally cheap, I know.  Everyone's gold income would have to be capped or scaled according to availability, and availability would be a calculation based upon the town's percentage of world population, the world-total military/diplomatic upkeep, and the (possibly negative) global stockpile size when not all gold is in circulation.  "All gold" would itself be a periodically changing number based on population, and therefore the stockpile amount could be negative when population drops and gold in circulation is already over-extended.

To be a little more specific with a very simple (and of course untested) model:
wpop = world's aggregate population
wgold = sum of all saved gold in the world
wkeep = world's aggregate upkeep // should be updated at least hourly, and cascade to updating all town incomes
tcut = town's population / wpop
stockpile = 100 * wpop - wgold // could be just updated daily, can be negative, which means negative inflation
town income = min(tpop * taxation * wkeep, current income calculation) + tpop * stockpile

In theory the above will put the gold in circulation through a negative feedback loop trending toward the population-based limit, and at the most expense to players with the highest spending (upkeep).  It could instead cause the stockpile size to flip daily (or however often the stockpile is recalculated) by ever larger numbers causing wild fluctuation and break the whole system, or have some other balance issue.  But if so, it's just a matter of finding some altered version that maintains the balance properly (and could be as simple as making world upkeep and world stockpile always update at the same time).  This formula only serves as a theoretical starting point.

At any rate, such a system properly tuned would create a much more realistic economic balance where inflation is bound to the actual growth of the game, and the ability to keep gold will no longer be more valuable than its usefulness in trade.  The amount of gold in circulation could even be manually controlled based on market analysis by the devs to help maintain that balance and foster trade.  If they were to publish the total amount of gold, or tie it automatically to population, players chart out the actual global inflation themselves, lend their expertise as economists to the devs, and start more advanced financial operations such as investing in shares/stock and venture capital.

----

Once the currency problem is fixed, many other noted problems would solve themselves within the existing mechanics.  Gold would more meaningfully limit war operations without rich backing by non-militaristic support.  The herald could start publishing meaningful economic information based upon the "daily close" of trading and state of the new metrics being used/managed.  People who wish to run shops could set up regular supply lines, invest in futures, sell stock/shares to fund startup, etc.  As a direct result of the increase in middle men, the "insurmountable" distance problem would dissolve.  It would then be prudent to allow offering "perpetual trade offers," by which I mean a trader puts up an offer which when accepted is automatically replaced with another identical offer so long as the caravans and resources being sold are available.  When the potential profit in running trade operations becomes realizable, the people who then devote themselves to trade will finally gain enough experience to start establishing market pricing.  When gold starts to hold its value, it becomes worthwhile to run banks and offer loans for wars or venture capital for startups.  People will start trying to stockpile gold for something other than saving it for a rainy day (the eventual short-term growth of armies to upkeep well beyond reasonable taxation).

Fix the currency, and trade will become the rich and rewarding facet of the game as it should be, finally able to rival war as a meaningful first-class mechanic within the game.