How can I add columns to a feature collection in Google Earth Engine?

Question

I am trying to calculate the average nighttime radiance of forested areas in South Asia and Southeast Asia, broken up by country and year, then export the results as a single table where each row is a different country and each column is a different year. I have a program that will make a separate feature collection for each year, but I need some way to combine them all together as different columns of one table.

Currently, this code creates 8 different export tasks (1 for each year). Each task produces a table with 16 rows (1 for each country) and a single column. I would like it to instead create a single task, which would produce a table with 16 rows and 8 columns.

var asia = ["Sri Lanka","Bangladesh","India","Afghanistan",
            "Pakistan","Bhutan","Nepal","Cambodia","Indonesia", "Burma",
            "Laos","Malaysia","Philippines","Singapore","Thailand","Vietnam"];
        
var countries = ee.FeatureCollection("USDOS/LSIB/2017");

var asiafeatures = asia.map(function(element){  // feature collection of country names
  return ee.Feature(null, {'country': element});
})

// initialize table
var table = ee.FeatureCollection(asiafeatures);

var years = ee.List.sequence(2014, 2021); //  list of numbers 2014 to 2021

var fields = years.map(function (number) {
  return ee.String(number).slice(0, 4); //  count only pixels with values of 0, 1, 2, or 3
})

var size = years.length()

var covers = ee.ImageCollection.fromImages( //  assemble an image collection from images
  years.map(function (y) {  //  for each year
            return ee.ImageCollection("MODIS/061/MCD12Q1")  //  from the modis land cover image
              .select('LC_Type1') //  select this field
              .filter(ee.Filter.calendarRange(y, y, 'year'))  //  filter it by the desired timespan
              .median() //  find the median
              .set('year',y); //  and set the value of the year field?
    }).flatten());  //  flatten the MODIS collection to 1 image

var rads = ee.ImageCollection.fromImages( //  assemble an image collection from images
  years.map(function (y) {  //  for each year
            return ee.ImageCollection("NOAA/VIIRS/DNB/MONTHLY_V1/VCMSLCFG") //  from the nightlight image
              .select('avg_rad')  //  select this field
              .filter(ee.Filter.calendarRange(y, y, 'year'))  //  filter by the desired timespan
              .median() //  find the median
              .set('year',y);
    }).flatten());

var coverslist = covers.toList(size); //  turn the land cover image collection into a list of images, 1 per year
var radslist = rads.toList(size); //  turn the nightlights image into a list of images, 1 per year
var sequence = ee.List.sequence(0, ee.Number(size).subtract(1));  //  list of numbers from 0 to 7

var stats = sequence.map(function(seq){ // for each number in the sequence (each year)
  
  return ee.FeatureCollection(asia.map(function(c){ //  for each country in asia
    var country = countries.filter(ee.Filter.eq('COUNTRY_NA', c));  // one country at a time
    //  country is a feature collection with 1 feature
    var coverclipped = ee.Image(coverslist.get(seq)).clip(country); //  grab an image from the list, serve it, clip
    //  you can't print this one because it's server-side, but it's an image
    var coverclippedmasked = coverclipped.updateMask(coverclipped.lte(6));  //  clip it to forest areas?
    //  still an image, still server-side
    var binarymask = coverclippedmasked.selfMask().mask().clip(country).toInt();  //  simplify to forest and not?
    //  yep
    var vectors = binarymask.reduceToVectors({  //  turn the mask into a polygon
        geometry: country,  //  separated by country?
        scale: 500,
        maxPixels: 1e18
    });
    //  vectors is a feature collection
    var forestRadianceMed = ee.Image(radslist.get(seq)).reduceRegions({ //  reduce the clipped nightlights by region
        collection: vectors.filter(ee.Filter.eq('label', 1)), // forested polygons only
        reducer: ee.Reducer.median(),
        scale: 500
    });
    // forestRadianceMed is a feature collection
    return ee.Feature(null, {}).set(ee.String(fields.get(seq)), forestRadianceMed.aggregate_mean('median'))
    //  this is a feature
  }));
});

// this iterates through the list of computed objects and exports each collection individually

for(var i = 0; i < size.getInfo(); i++) {
  Export.table.toDrive({
    collection: stats.get(i),
    description: years.get(i).getInfo() + '_forestRadianceMed',
    selectors: [years.get(i).getInfo()]
  })
  
}

// Well, it does the calculations, but they're still not all in one table.